模糊神经网络控制器在ROV机械手上的应用研究  

Application of Fuzzy Neural Network Controller in ROV's Mechanical Hands

作　　者：田烈余[1,2] 盛堰[1,2] 陈春亮[1,2] 

机构地区：[1]广州海洋地质调查局,广东广州510760 [2]国土资源部海底矿产资源重点实验室,广东广州510760

出　　处：《海洋技术》2013年第4期93-95,共3页Ocean Technology

基　　金：国家高技术研究发展计划(863计划)资助项目(GZH201100307-04-01)

摘　　要：针对水下机器人机械手抓取专用工具及操作准确、快速、可靠平稳的要求,设计一种应用ROV的模糊神经网络控制器。该控制器将PID控制、模糊控制和神经网络相结合,设计一种基于模糊神经网络(FNN)的复合控制器,并将其应用于机械手的位置控制中,研究变轨迹变周期情况下机械手的位移跟踪特性。仿真结果表明:该控制器具有良好的动态、稳定性能以及较强的鲁棒性,能够使水下机器人的机械手操作快速准确平稳。To meet the design requirements of underwater robots, a fuzzy neural network controller equipped with ROV was designed, enabling the mechanical hands to snatch special tools in an accurate, fast, stable and reliable way. Integrated with PID control, fuzzy control and neural network, the controller realized a compound controller based on Fuzzy Neural Network （FNN）, which was applied in the control and positioning of mechanical hands. The displacement tracking features of mechanical hands were studied in the conditions of changeable trajectory and cycle. The simulation results show that the controller has good dynamic and stable performance and strong robustness, which enables the mechanical hands of underwater robots to operate in a fast, accurate and smooth way.

关 键 词：ROV 模糊 神经网络 仿真 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]