文章摘要
基于正交神经网络的动力定位自适应控制器设计
Dynamic Positioning Adaptive Controller Design Based on Orthogonal Neural Network
投稿时间:2018-05-22  修订日期:2018-07-25
DOI:
中文关键词: 动力定位,定点定位,自适应控制,模型试验
英文关键词: dynamic positioning, station keeping, adaptive control, model experiment
基金项目:国家自然科学基金 (项目批准号:61301279,51479158);中央高校基本科研业务费专项资金资助(项目批准号:172102003);高性能船舶技术教育部重点实验室开放基金课题(项目批准号:2016gxn01)
作者单位E-mail
徐海祥 武汉理工大学 qukaiyang@163.com 
摘要点击次数: 24
全文下载次数: 
中文摘要:
      动力定位船舶在海上进行定点定位作业时不可避免地会受到风、浪、流海洋环境力的干扰,这种持续的环境力扰动会导致船舶在实际定位时产生静态误差,不能准确地到达目标定位点。在利用反步方法进行控制器设计时,大多数文献引入自适应积分项用于抵抗环境载荷干扰,通过在控制器中加入船舶当前位置与目标位置的偏差积分项估计出外界未知环境扰动,从而达到自适应的效果。本文在此基础上,改进了自适应积分项为正交基神经网络项进行控制器设计,以达到消除静态偏差,实现准确定位的目的。最后通过水池模型试验的对比验证了所设计控制器的可行性。
英文摘要:
      Dynamic positioning ship will inevitably be disturbed by marine environmental forces such as wind, waves and current which causing static errors in the actual positioning of the ship. When using the backstepping method to design the controller, most of the documents introduce adaptive integral to resist the environmental load interference. By adding the deviation integral of the current position of the ship and the target position in the controller, the unknown environment disturbance is estimated, thus the adaptive effect is achieved. This paper improves the adaptive integral term for the orthogonal basis neural network to design the controller, and achieves the purpose of eliminating static deviation and realizing the accurate positioning. Finally, the feasibility of the designed controller is verified through the comparison of model experiments.
View Fulltext   查看/发表评论  下载PDF阅读器
关闭