基于人工蜂群优化的高超声速飞行器在线参数辨识  

Artificial bee colony approach to online parameters identification for hypersonic vehicle

作　　者：李霜天[1] 段海滨[1,2] 

机构地区：[1]北京航空航天大学自动化科学与电气工程学院、飞行器控制一体化技术重点实验室,北京100191 [2]苏州大学江苏省计算机信息处理技术重点实验室,苏州215006

出　　处：《中国科学：信息科学》2012年第11期1350-1363,共14页Scientia Sinica(Informationis)

基　　金：国家自然科学基金(批准号:61273054,60975072);国家高技术研究发展计划(863计划)(批准号:2011AA040902);航空科学基金(批准号:20115151019);教育部新世纪优秀人才支持计划(批准号:NCET-10-0021);苏州大学江苏省计算机信息处理技术重点实验室开放课题项目基金(批准号:KJS1020)资助项目

摘　　要：高超声速飞行器由于具有特殊的气动特性和复杂的运行环境,其气动模型的建立和模型中参数的确定面临着更高的要求.飞行器参数辨识是根据飞行器的输入及其响应确定出飞行器的模型和模型中的各个参数数值.针对高超声速飞行器模型耦合性强、非线性程度高、运行环境复杂等特点,本文提出了基于人工蜂群优化的在线参数辨识方法,将参数辨识问题转换为优化问题,以蜂群为单位进行搜索,通过群体信息交流和优胜劣汰的机制,使得蜂群朝着更优方向进化;引入采蜜蜂机制和混沌搜索机制,使得蜂群能够跳出局部最优,具有更强的全局寻优能力.应用此方法对某飞行器升力系数进行辨识计算,结果证明了此方法的可行性.与传统的极大似然法对比表明,本文所提方法在具有系统测量噪声的条件下具有更强的抗干扰能力和准确性.Due to the special aerodynamic characteristics and complexity of operating environment, the modeling and determination of the model parameters are faced with higher requirements. The employment of aircraft parameter identification is aimed to determine the model and model＇s parameters of an aircraft by observing the aircraft＇s inputs and responses. In this paper, an artificial bee colony based approach is proposed for solving the parameter identification problem under the circumstance of high level of model coupling, nonlinearity and complex operating environment of hypersonic vehicles. Firstly, the parameter identification problem is transformed into an optimization problem, such that feasible solutions of the problem can be denoted as food sources for which the artificial bee colony is seeking. Then information exchanges among the colony and the survival of the fittest are introduced to enable the artificial bee colony to evolve toward a better direction. Scouting bees and chaotic search strategy are adopted, which enable the swarm to jump out of local optima. The feasibility of the proposed method is proved by its implement to identify the lift coefficient of a hypersonic vehicle. A series of comparative results with maximum likelihood method show the better performance of the presented method in systems with measurement noises.

关 键 词：高超声速飞行器 参数辨识 人工智能 人工蜂群优化 极大似然法 全局优化 

分 类 号：V221.3[航空宇航科学与技术—飞行器设计]