Aero-engine interpolation model modeling method based on machine learning

Abstract

The invention discloses an aero-engine interpolation model modeling method based on machine learning, and the method comprises the following steps: collecting flight data and corresponding key parameters of an aero-engine, carrying out the feature selection, carrying out the preprocessing of original data, and building a flight data set; taking the converted input features as node coordinates, taking the converted output features as target values, performing retrieval in a flight data set, preliminarily establishing an interpolation table, and marking missing nodes in the interpolation table; using samples in the flight data set to establish a machine learning model using the reduced features as input and output; and predicting missing nodes of the interpolation table by using a machine learning model to obtain a complete interpolation table. According to the method, the problem that the precision is reduced due to the fact that a traditional interpolation model is limited by the working condition range of flight data is solved, the powerful generalization performance of the random forest is utilized, the complete and accurate interpolation table is established, and the precision of the interpolation model of the aero-engine can be effectively improved in the actual use process.

Description

technical field [0001] The invention relates to the modeling of key parameters of aero-engine, in particular to a modeling method of aero-engine interpolation model based on machine learning. Background technique [0002] With the continuous improvement of modern aircraft performance and engine performance, various control methods emerge in an endless stream. Afterburner fuel control, anti-surge control, etc. bring more and more control variables, which put forward higher and higher requirements for the engine control system. Since the advent of the full authority digital electronic controller in 1970, it has quickly replaced the traditional hydraulic system due to its simple structure, light weight, good adaptability and high reliability, and has been widely used in civil and military aero-engines. To develop such a complex and advanced aero-engine numerical control system, from the initial control law analysis and design to the later semi-physical simulation test, it is ne...

AI Technical Summary

Problems solved by technology

At present, there is still a large research gap in the field of applying machine learning algorithms to the modeling of aero-engine interpolation models

Images