【作者】 胡涛；

【导师】 牟军敏； 陈琳瑛；

【作者基本信息】 武汉理工大学 ， 交通信息工程及控制， 2022， 硕士

【摘要】 水上搜救工作是保障海上人命安全的最后一道防线。高效精准的海上搜救是世界性的难题,研究智能搜救新装备和新技术,提高水上搜救能力,是建设交通强国的重要发展方向,对保障人民群众生命财产安全、保护海洋生态环境、服务国家发展战略、提升国际影响力、彰显社会主义制度优势具有重要作用。无人艇(Maritime Autonomous Surface Ship,MASS)在海上落水人员搜寻方面具有突出的优势,研究无人艇编队协同搜寻方法,可有力地支持海事搜救装备发展,有助于显著提高海事搜救的效率与水平,切实保障人民生命财产安全。本论文面向人员落水搜寻事件,提出基于无人艇编队的海上搜寻研究框架,结合落水人员海上漂移运动预测,规划无人艇编队协同搜寻轨迹。通过搜寻路径的时间与系统约束,采取集中-分布式结构,基于模型预测控制(Model Predictive Control,MPC)方法,实现群体自治机器人海上自主协同搜寻的轨迹跟踪控制研究。本文主要研究内容如下:(1)无人艇编队落水人员搜寻框架:针对人员落水事件设计无人艇编队搜寻轨迹规划和轨迹跟踪框架,具体包括位置预测、搜寻轨迹规划、轨迹跟踪三个功能模块。(2)无人艇编队协同搜寻轨迹规划算法:根据海洋环境与落水人员海上漂移运动特点,基于蒙特卡洛法随机粒子仿真预测落水人员漂移位置,建立搜寻区域,并提出自适应贪婪算法,动态规划无人艇编队海上协同搜寻轨迹。(3)无人艇编队协同搜寻轨迹跟踪算法:采取集中-分布式控制结构,基于集中式规划编队轨迹,确定编队内单艇搜寻轨迹,基于模型预测控制方法,设计分布式无人艇轨迹跟踪算法。最后通过仿真实验,验证本文提出的无人艇编队协同搜寻轨迹规划及跟踪算法在落水人员搜寻任务中的可行性及有效性。本研究主要创新点如下:(1)提出了一种改进的落水人员位置漂移轨迹预测模型。模型考虑海上人员落水时间不同场景、海洋环境条件,基于人员在船舶航行过程中落水初始概率分布模型,为搜寻区域的快速确定提供准确的数据。(2)提出了一种融合落水人员漂移位置动态预测的无人艇编队协同搜寻轨迹规划算法,大幅度提高海上遇险目标的搜寻效率。(3)提出了一种基于MPC方法的无人艇编队轨迹跟踪算法,兼顾无人艇编队搜寻灵活性与轨迹跟踪控制精度,有效实现协同搜寻任务。论文研究成果可为海上搜救中心应急搜寻、指挥调度等工作提供重要的理论和技术支持,有助于加快推进无人艇编队在海上搜救应用与发展,为提高海上搜救能力和落水人员搜寻成功率贡献力量。更多还原

【Abstract】 Search and Rescue(SAR)at sea is the last barrier of safety life at sea.Efficient and accurate patterns of maritime SAR are still challenging worldwide.Emerging intelligent SAR equipment and technology is the frontier of national power in transportation.It plays an important role in safeguarding people’s lives property and the marine ecological environment,which also serves the national development strategy and boosts China’s international clout.Maritime Autonomous Surface Ship(MASS)has outstanding advantages in SAR for man overboard(MOB)at sea,and it can strongly support the development of the maritime SAR industry,and significantly improve the efficiency and feasibility of successful SAR.This paper establishes a framework of MASS cooperative search for MOB at sea.Combined with the drift prediction of the missing person,the cooperative search trajectory of the MASS fleet is planned.According to the time and system constraints of the search path,a centralized-distributed control structure is applied for trajectory tracking of the MASS fleet based on the model predictive control(MPC)method.This paper focuses on the following work:First,a framework of MASS cooperative search for MOB at sea is established.Focusing on the man overboard incident,this paper proposes the trajectory planning and tracking framework of MASS formation,consisting of three functional modules,position prediction,maritime search trajectory planning,and trajectory tracking.Second,a cooperative MASS trajectory planning for marine man overboard search is developed.The stochastic particle simulation method is used to predict the missing person’s position considering the environment forecasting data,which determines the search area.An adaptive greedy trajectory planning algorithm is designed with an adaptive neighborhood and evaluation function for increasing the cumulative probability of success in a limited time.Third,a trajectory tracking algorithm is designed based on MPC method for the cooperative search.We designed a structure combining centralized trajectory planning and distributed trajectory control.The search trajectory of each MASS in the fleet is determined according to the above-mentioned trajectory for the cooperative SAR.The formation trajectory tracking is realized through each MASS tracking its own trajectory with an MPC controller.Simulation experiments are carried out to demonstrate the feasibility and effectiveness of the proposed framework of MASS cooperative search for MOB at sea.The main innovative idea of this paper are as follows:First,an improved drift prediction model of the MOB is presented.This model takes marine environment in different scenarios and the uncertainty of initial position into account,supporting rapid determination of search areaSecond,a cooperative MASS trajectory planning for marine man overboard search is developed.The algorithm integrates the dynamic prediction of drift position of MOB and cumulative probability of success in the evaluation function,which can greatly improve the search efficiency.Third,a formation track tracking algorithm based on MPC method for cooperative SAR is developed,which takes into account the flexibility of MASS formation search and the control accuracy of trajectory tracking.The research results of this paper can provide important theoretical and technical support for the maritime SAR.Besides,it will help to speed up the application and development of the MASS formation in maritime SAR,and contribute to improving maritime SAR abilities and the success rate of searching for MOB.更多还原