摘要
针对基于部件级航空发动机动态建模过程中完整、准确的航空发动机部件特性数据往往难以获取、建模时间长等现象,提出使用实验数据进行辨识建模的方法。为了建立航空发动机的动态模型,通过对某轻型飞机实验台的飞行实验数据进行分析整理,提出使用BP神经网络对发动机重要参数进行建模,同时使用粒子群优化算法(Particle swarm optimization,PSO)对BP神经网络的权值和阈值进行优化,使用改进粒子群优化算法(Improved particle swarm optimization algorithm,IPSO)对传统粒子群优化算法进行改进,仿真结果表明IPSO-BP网络建立的发动机模型精度和稳定性更高。
It is difficult to obtain the complete and accurate aeroengine component characteristic data in the dynamic modeling process of component-based aeroengine and the modeling time is long.So the method of using experimental data for identification modeling is proposed.In order to establish the dynamic model of the aeroengine,by analyzing the flight experimental data of a light aircraft test bench,it is proposed to use BP neural network to model the important parameters of the engine.At the same time,particle swarm optimization(PSO)is used to optimize the weight and threshold of BP neural network.Finally,the improved particle swarm optimization algorithm(IPSO)is used to improve the traditional particle swarm optimization algorithm.The simulation results show that the engine model established by IPSO-BP network has higher precision and better stability.
作者
马振
张正
刘程
MA Zhen;ZHANG Zheng;LIU Cheng(Civil Aircraft Airworthiness Certification Technology Key Laboratory,Civil Aviation University of China,Tianjin 300300,China)
出处
《滨州学院学报》
2019年第4期22-28,共7页
Journal of Binzhou University
关键词
航空发动机
模型辨识
动态建模
神经网络
改进粒子群优化算法
aeroengine
model identification
dynamic modeling
neural network
improved particle swarm optimization
