基于CARMA模型的水下滑翔机自适应控制  

CARMA Model Based Adaptive Controller For Underwater Glider

作　　者：赵宝强[1] 王晓浩[1] 姚宝恒[2] 连琏[1,2] 

机构地区：[1]清华大学精密仪器系,北京100084 [2]上海交通大学海洋研究院,上海200240

出　　处：《计算机仿真》2014年第9期374-378,共5页Computer Simulation

基　　金：国家科技重大专项(2011ZX05027);国家自然科学基金(51279107);上海市科研计划项目(13dz1204600)

摘　　要：水下滑翔机在设计和海中实际应用过程中存在模型参数的不确定性和外部环境的复杂性,要求控制器具有良好的自适应性,在水下滑翔机三维定常运动模型基础上,以俯仰回路模型为例设计了基于CARMA模型的改进广义预测自适应控制器。上述控制器应用遗忘因子递推最小二乘法对系统参数进行实时辨识。在对控制器进行了仿真后,对CARMA模型参数全部突变这一恶劣情况进行仿真并与PID控制对比,结果表明改进控制算法具有良好的快速性、稳定性和自适应性。Considering the model parameters uncertainty and complexity of the external environment in the design and practical application of the underwater glider, the controller of glider should be adaptable. So an improved generalized predictive adaptive controller based on CARMA model was designed for pitch loop model based on three-dimensional model of glider. Forgetting factor recursive least squares method was used to identify the real-time parame- ters. After simulation of the controller, a worst situation that modal parameters varied sharply was considered and simulated. Results show that the control algorithm has good speed, stability and adaptability.

关 键 词：水下滑翔机 广义预测控制 递推最小二乘法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]