多模自适应控制在空中飞行模拟器中应用  

Application of Multi-model Adaptive Control to In-flight Simulator

作　　者：刘兴堂[1] 王革命[1] 张双选[1] 

机构地区：[1]空军工程大学导弹学院,三原713800

出　　处：《系统仿真学报》2003年第1期129-131,共3页Journal of System Simulation

基　　金：国防基金航空工程仿真新理论新方法研究(40403201JB201)

摘　　要：为了克服空中飞行模拟器传统上使用增益调度控制方式的弱点,设计了一种改善空中飞行模拟器性能的自适应模型跟随控制系统。其中辨识器采用基于卡尔曼滤波器的多模辨识方法,该方法将本机先验信息加入到参数辨识过程中,辨识结果是与卡尔曼滤波器相对应的各个先验模型参数的概率权重和。仿真结果表明,该控制系统克服了原空中飞行模拟器模拟精度依赖于需要精确已知本机气动导数的严重缺陷,并显示出多模辨识算法比一般的辨识方法收敛速度快、精度高,有效地解决了非线性随机系统的参数辨识问题。To overcome the weakness of using the gain scheduling control method in In-flight simulator, an adaptive model following control system is developed to improve the performance of in-flight simulator. A multi-model identification algorithm based on Kalman filter is used in the identificator. By incorporating a-priori information of the host into the identification process, the identified parameters are formed as a probabilistic weighted sum of individual parameters of a-priori models. The simulation results show that the control system overcomes the shortcoming of the old in-flight simulator whose simulation precision relies on precise air dynamics derivative of the host plane, and multi-model identification algorithm is superior to traditional identification method in convergence speed and accuracy, and it is efficient to identify the parameters of the no-linear probabilistic systems.

关 键 词：多模自适应控制 空中飞行模拟器 参数辨识 卡尔曼滤波 应用 

分 类 号：V216.7[航空宇航科学与技术—航空宇航推进理论与工程] TP273.2[自动化与计算机技术—检测技术与自动化装置]