基于LPV的航空发动机压气机健康监测系统建模研究  

Research on Aero-Engine Compressor Modeling for Health Monitoring System Based on LPV

作　　者：贺琛[1] 张小栋[1] 

机构地区：[1]西北工业大学动力与能源学院,陕西西安710072

出　　处：《测控技术》2014年第5期139-142,共4页Measurement & Control Technology

基　　金：国家公派出国留学资助项目

摘　　要：针对航空发动机健康监测系统对模型的高实时性和准确性要求,提出了一种应用递归最小二乘线性拟合建立压气机非线性过程中的线性变参数(LPV)模型的建模方法。该建模方法通过对模型参数的选择和顶点的选取,结合递归最小二乘法在不同顶点处拟合线性状态空间模型,根据LPV模型凸集特点得到非线性模型的LPV模型表达。最后应用法国ALSTOM公司提供的燃气涡轮模型进行了仿真实验验证,实现了实时非线性过程中模型稳定区域小于1%的动态建模误差,证明了方法的有效性和精确性。Due to high real-time and accuracy requirements of the model of aero-engine health monitoring system,a linear parameter varying （LPV） methodology is proposed to establish the model of the compressor nonlinear process by using the recursive least square linear fitting method.The modeling method can be divided as follows：parameters and vertex selection linear state space modeling by applying recursive least square method in different vertex LPV model representation of the nonlinear process according to the characteristics of convex sets.The performance of the modeling method is demonstrated by applying it to the nonlinear gas turbine compressor model provided by ALSTOM company,and the result shows that less than 1％ dynamic modeling error can be generated which verifies the validity and accuracy of the LPV method.

关 键 词：航空发动机健康监测 压气机模型 线性变参数模型 递归最小二乘法 

分 类 号：V249[航空宇航科学与技术—飞行器设计]