带有模型降阶的μ综合算法及应用  

A New Method of μ Synthesis with Model Reduction and Its Application

作　　者：阮洪宁[1] 史忠科[1] 

机构地区：[1]西北工业大学

出　　处：《飞行力学》1999年第2期76-81,共6页Flight Dynamics

基　　金：国家自然科学基金;国防基金

摘　　要：针对在μ综合中产生的控制器阶次高、可实现性差等实际问题，提出了一种带有平衡截断法的μ综合算法。该算法原理是通过可控可观测Ｇｒａｍｍｉａｎ阵的折中，达到降阶且保持原系统性能不变的目的。新方法不仅非常接近最优结果，而且可以使控制器阶次大大降低。对飞机纵向运动μ综合仿真结果表明，该降阶方法可以使控制器阶次由２０阶降为５阶，而且工程上可以实现。To solve the problems of both high order and numerical“ill conditioned” in the μ controller obtained by the D K iteration algorithm,a new approach of μ synthesis with balanced realization is proposed. In the new approach,a balanced minimal realization is obtained by trade off the degree between Grammian matrices of controllability and observability of different subspaces of the state space. Thus a low order μ controller is obtained and the performance of the new model remains approximate to the original system. By means of the new approach, the controller is not only very near to the optimal ones but also the order is decreased greatly. Simulation results of μ synthesis to aircraft longitidunal motion is excellent and the order of μ controller is decreases from 20 dimensional system to 5 ones.

关 键 词：μ-控制器 模型降阶 平衡实现 综合算法 飞机 

分 类 号：V249.121[航空宇航科学与技术—飞行器设计]