【作者】 施阳； 慕春棣； 严卫生； 李俊； 徐德民；

【Author】 Shi Yang, Mu Chundi Institute of Control Theory and Technology Department of Automation, Tsinghua University Beijing, 100084 Yan Weisheng, Li Jun, Xu Demin Institute of AUV Northwestern Polytechnical University, Xi’an, 710072

【机构】 清华大学自动化系控制理论与技术研究所； 西北工业大学水下自航器研究所；

【摘要】 在研究多层前向网络逼近性质的基础上,提出一种基于多层前向网络的稳定自适应控制方案,并应用于水下自航器纵向控制系统的设计。该方案采用多层前向网络在线逼近复杂不确定性非线性,闭环控制系统中参数的的有界性及跟踪误差的收敛性采用Lyapunov稳定性理论进行了证明。仿真结果验证了该方案的有效性。更多还原

【Abstract】 Most existing AUV control schemes are based on its simplified invariant linear mathematical model. However, AUV has complex uncertain nonlinearities. We succeeded in proposing a stable adaptive neural-network-based control scheme for AUV’s depth control and then achieving global tracking convergence. Multi-layer feedforward neural networks with sigmoid hidden units are analyzed in detail for controller design(Theorem 1). The control law we designed for AUV’s depth system is eqs. (13). To go along well with eqs. (13), we determined eqs. (15) through (18) as the adaptive laws of neural network’s weights and other parameters. Then, utilizing Lyapunov stability theorem and Barbalat Theorem, we proved that the tracking error converges to a neighborhood of zero and all parameters in the closed-loop system are all bounded. Simulation results(Fig.2 ) show that the scheme we proposed is effective.更多还原