Path Following Control of A Deep-Sea Manned Submersible Based upon NTSM

In this paper, a robust path following control law is proposed for a deep-sea manned submersible maneuvering along a predeterminated path. Developed in China, the submersible is underactuated in the horizontal plane in that it is actuated by two perpendicular thrusts in this plane. The advanced non-singular terminal sliding mode （NTSM） is implemented for the design of the path following controller, which can ensure the convergence of the motion system in finite time and improve its robustness against parametric uncertainties and environmental disturbances. In the process of controller design, the close-loop stability is considered and proved by Lyapunov＇ s stability theory. With the experimental data, numerical simulations are provided to verify the control law for path following of the deep-sea manned submersible.     

Adaptive Predictive Inverse Control of Offshore Jacket Platform Based on Rough Neural Network

The offshore jacket platform is a complex and time-varying nonlinear system,which can be excited of harmful vibration by external loads.It is difficult to obtain an ideal control performance for passive control methods or traditional active control methods based on accurate mathematic model.In this paper,an adaptive inverse control method is proposed on the basis of novel rough neural networks (RNN) to control the harmful vibration of the offshore jacket platform,and the offshore jacket platform model is established by dynamic stiffness matrix (DSM) method.Benefited from the nonlinear processing ability of the neural networks and data interpretation ability of the rough set theory,RNN is utilized to identify the predictive inverse model of the offshore jacket platform system.Then the identified model is used as the adaptive predictive inverse controller to control the harmful vibration caused by wave and wind loads,and to deal with the delay problem caused by signal transmission in the control process.The numerical results show that the constructed novel RNN has advantages such as clear structure,fast training speed and strong error-tolerance ability,and the proposed method based on RNN can effectively control the harmfid vibration of the offshore jacket platform.     

Neural Network Based Robust Controller for Trajectory Tracking of Underwater Vehicles

A robust neural network controller (NNC) is presented for tracking control of underwater vehicles with uncertainties. The controller is obtained by using backstepping technique and Lyapunov function design in combination with neural network identification. Modeling errors and environmental disturbances are considered in the mathematical model. A two-layer neural network is introduced to compensate the modeling errors, while H∞ control strategy is used to achieve the L2-gain performance. The uniformly ultimately bounded (UUB) stabilities of tracking errors and NN weights are guaranteed through the proposed controller. An on-line NN weights tuning algorithm is also proposed. Good performances of the tracking control system are illustrated by the results of numerical simulations.     

Experimental Study of a Command Governor Adaptive Depth Controller for an Unmanned Underwater Vehicle

Unmanned Underwater Vehicles (UUVs) are increasingly being used in advanced applications that require them to operate in tandem with human divers and around underwater infrastructure and other vehicles. These applications require precise control of the UUVs which is challenging due to the non-linear and time varying nature of the hydrodynamic forces, presence of external disturbances, uncertainties and unexpected changes that can occur within the UUV’s operating environment. Adaptive control has been identified as a promising solution to achieve desired control within such dynamic environments. Nevertheless, adaptive control in its basic form, such as Model Reference Adaptive Control (MRAC) has a trade-off between the adaptation rate and transient performance. Even though, higher adaptation rates produce better performance they can lead to instabilities and actuator fatigue due to high frequency oscillations in the control signal. Command Governor Adaptive Control (CGAC) is a possible solution to achieve better transient performance at low adaptation rates. In this study CGAC has been experimentally validated for depth control of a UUV, which is a unique challenge due to the unavailability of full state measurement and a greater thrust requirement. These in turn leads to additional noise from state estimation, time-delays from input noise filters, higher energy expenditure and susceptibility to saturation. Experimental results show that CGAC is more robust against noise and time-delays and has lower energy expenditure and thruster saturation. In addition, CGAC offers better tracking, disturbance rejection and tolerance to partial thruster failure compared to the MRAC.     

浅谈"蛟龙"号载人潜水器试验性应用航次陆基保障工作

陆基保障工作是依托“蛟龙”号载人潜水器(以下简称“蛟龙”号)试验性应用航次开展的一项兼具日常公务办理和应急事件处置的综合事务性工作,主要工作内容包括：跟踪航次动向、保障海陆联通、协调船舶靠港、落实航次宣传以及为“蛟龙”号试验性应用工作领导小组和海上现场指挥部提供技术支撑等。文章回顾了“蛟龙”号试验性应用航次陆基保障工作开展情况,分析了工作特点和存在问题,针对下一步多机构联合和多潜水器协同作业模式下的深远海科考航次特点,提出了陆基保障工作的数字化建设和保密机制建设初步设想。     

基于模糊神经网络理论对水下拖曳体进行深度轨迹控制

以华南理工大学开发的自主稳定可控制水下拖曳体为研究对象,首先通过水下拖曳体在拖曳水池样机中的试验取得试验数据后作为训练样本,采用LM BP算法,建立基于神经网络理论构建的可控制水下拖曳体轨迹与姿态水动力的数值模型。在此基础上设计了一个控制系统,它主要由两部分组成:基于遗传算法的神经网络辨识器和基于模拟退火改进的遗传算法的模糊神经网络控制器。以满足预先设定的拖曳体水下监测轨迹要求为控制依据,由控制系统确定为达到所要求的运动轨迹而应采用的迫沉水翼转角,以此作为输入参数,通过LM BP神经网络模型的模拟计算预报在这一操纵动作控制下的拖曳体所表现的轨迹与姿态特征。数值模拟计算结果表明:该系统的设计达到了所要求的目的;借助这一系统,可以有效地实现对拖曳体的深度轨迹控制。     

一种基于BP算法学习的小波神经网络

为发展 Szu的基于信号表示的小波神经网络 ,提出一种多输入多输出的小波网络模型 ,网络隐层采用框架小波函数、输出层采用 Sigmoid激励函数 ,并选用“熵误差函数”以加速网络的学习速度。奇偶判别和混沌时间序列预测例子的实验结果表明了它具有良好的函数逼近能力和推广能力 ,收敛速度和均方误差均优于相同结构的多层感知器模型。     

水下自主机器人水动力预报方法研究

针对水下机器人操纵性优化设计中水动力系数预报问题,在水下机器人水动力预报中引入艇体肥瘦指数概念,确定了水下机器人艇体几何描述的五参数模型。提出采用小波神经网络方法预报水下机器人水动力,确定了神经网络的结构,利用均匀试验设计方法,设计了神经网络的学习样本。研究结果表明,只要确定适当的输入参数,选择适当的学习样本和网络结构,利用小波神经网络方法对水下机器人水动力进行预报可以达到较好的精度。     

Parallel Neural Network-Based Motion Controller for Autonomous Underwater Vehicles

A parallel neural network-based controller （PNNC） is presented for the motion control of underwater vehicles in this paper. It consists of a real-time part, a self-learning part and a desired-state programmer, and it is different from normal adaptive neural network controller in structure. Owing to the introduction of the self-learning part, on-line learning can be performed without sample data in several sample periods, resulting in high learning speed of the controller and good control performance. The desired-state programmer is utilized to obtain better learning samples of the neural network to keep the stability of the controller. The developed controller is applied to the 4-degree of freedom control of the AUV “IUV- IV” and is successful on the simulation platform. The control performance is also compared with that of neural network controller with different structures such as normal adaptive neural network and different learning methods. Current effects and surge velocity control are also included to demonstrate the controller＇ s performance. It is shown that the PNNC has a great possibility to solve the problems in the control system design of underwater vehicles.     

随机系统稳健自适应控制

对具有未建模动态特性且时滞任意的多维随机系统,通过采用有界外来激励和随机变界截尾方法,本文建立了推广最小二乘（ＥＬＳ）算法和加权ＥＬＳ算法的稳健估计,进而得到间接自适应模型参考控制的稳健性     

The Depth Limits of Red Sea Stony Corals: An Ecophysiological Problem (A Deep Diving Survey by Submersible)

Abstract. Along the Sinai coast of the Red Sea a deep diving survey with the research submersible GEO investigated the depth distribution of stony corals. 47 hermatypic species were identified below 50 m; 9 species extended below 100 m. Their depth distributions are related to light penetration. Observed changes in hermatype growth forms with depth were investigated and interpreted as photo-adaptations. 10 species of ahermatypic corals were found between 100–205 m. Ahermatype growth forms are adaptations to plankton feeding and do not change with depth.     

A chattering-free sliding-mode controller for underwater vehicles with fault-tolerant infinity-norm thrust allocation

There are two objectives to this paper. First, a chattering-free sliding-mode controller is proposed for the trajectory control of remotely operated vehicles (ROVs). Second, a new approach for thrust allocation is proposed that is based on minimizing the largest individual component of the thrust manifold. With regards to the former, a new adaptive term is developed that eliminates the high-frequency control action inherent in a conventional sliding-mode controller. As opposed to the common adaptive approach, the new adaptive term does not require the linearity condition on the dynamic parameters and the creation of a regressor matrix. In addition, it removes the need for a priori knowledge of upper bounds on uncertainties in the dynamic parameters of the ROV. With regards to the latter, it is demonstrated that minimizing the l_∞ norm (infinity-norm) of the thrust manifold ensures low individual thruster forces. The new control and thrust allocation concepts are implemented in numerical simulations of a work class ROV, and the chattering-free nature of the controller is demonstrated during typical ROV manoeuvres. In the simulation studies, the l_∞ norm-based thrust allocation problem is cast as a linear programming problem that allows direct incorporation of the thruster saturation limits and a fault-tolerant property. To achieve real-time solution rates for the l_∞ norm-based thrust allocation problem, a recurrent neural network is designed. In the simulation studies, the l_∞ norm-based thrust allocation provides smaller maximum absolute value of the largest component of the thrust manifold than that of a conventional l₂ norm (2-norm) minimization, satisfies the saturation limits of each thruster, and accommodates faults that are introduced arbitrarily during the manoeuvre.     

Neural Network-Based Active Control for Offshore Platforms

1 .IntroductionWiththedevelopmentofoceantechnology ,moreandmoreextremelylargeandlongflexibleoff shoreplatformsusedforoilexplorationanddrillingoperationarebuiltinhostileoceanenvironments .Ingeneral,thiskindofplatformsisanonlineardistributedparametersystemanditsnaturalfrequencyfallsclosertothedominantwavefrequencieswiththeincreaseofwaterdepth .Besides ,itsstructureisverycomplexandtheexternalwaveforceontheplatformisuncertain .Thus ,theseplatformsarepronetoexcessivewave inducedoscillationsunderbot…     

Coastal Wave Height Prediction using Recurrent Neural Networks (RNNs) in the South Caspian Sea

The prediction of wave parameters has a great significance in the coastal and offshore engineering. For this purpose, several models and approaches have been proposed to predict wave parameters, such as empirical, soft computing, and numerical based approaches. Recently, soft computing techniques such as recurrent neural networks (RNN) have been used to develop sea wave prediction models. In this study, the RNN for wave prediction based on the data gathered and the measurement of the sea waves in the Caspian Sea, in the north of Iran is used for this study. The efficiency of RNNs for 3, 6, and 12 hourly and diurnal wave prediction using correlation coefficients is calculated to be 0.96, 0.90, 0.87, and 0.73, respectively. This indicates that wave prediction by using RNNs yields better results than the previous neural network approaches.     

赤潮预测的人工神经网络方法初步研究

赤潮是一种由多因素综合作用引发的生态异常现象,具有突发性及非线性等特点。对其进行预测预报一直是海洋科学研究的热点。探讨了应用人工神经网络原理进行赤潮预测的方法,简要介绍了BP和RBF算法的基本原理,用2种算法对不同海域赤潮生物与环境因子之间非线性和不确定性的复杂关系进行学习训练和预测检验,并与传统的统计方法进行了比较。结果表明：人工神经网络方法在模拟和预测方面优于传统的统计回归模型,具有较强的模拟预测能力及实用性,值得进一步探索。     

水下机器人神经网络滤波技术的研究

滤波技术能提高水下机器人传感器信息的可靠性和精度 ,以保障水下机器人智能作业的顺利完成。将水下机器人的动力学特性隐式地分布在网络权值上 ,构造BP神经网络 ,进行水下机器人的运动状态预报 ,并将预报值与平滑后的实测数据相结合进行滤波 ,有效去除信号的噪声。仿真结果表明该方法能达到很好的滤波效果     

深海海底光缆敷设施工余量控制的原理和控制软件的应用

为解决深海有中继海底光缆项目技术难题,中国海底电缆建设有限公司开发了海底光缆敷设施工控制软件,填补我国相关领域的空白。文章介绍海底光缆敷设施工余量控制的原理和控制软件的操作流程:海底光缆余量包括区域余量、底部余量和释放余量,余量控制是海底光缆敷设施工中最关键的核心技术环节,应用控制软件可极大地降低计算量和提高计算精确度,并可通过在施工中不断调整计划,从而极大地提高施工质量,具有传统人工计算不可比拟的优势。     

海底观测网深水设备精准定点布放方法研究

海底观测网因其实时、长期、连续、高精度时钟同步及原位等优势而逐渐成为人类研究海洋的新型平台,建设规模和应用水深都在不断扩大。海底观测网系统建设中,深水设备的精准定点布放及湿插拔作业是施工的难点。针对国内海底观测网精准定位布放作业存在的困难和问题,结合国内现有施工条件,提出一种大深度海底设备精准定点布放安装方法,实现南海深海海底观测网试验系统深水设备精准定位布放与ROV湿插拔作业,对未来大规模海底观测网及其它深水工程中设备的精准定点布放和安装,具有参考和借鉴意义。     

卷积神经网络识别侧扫声呐影像的研究

卷积神经网络是当前应用最广泛的图像识别算法,利用大量数据对网络进行训练,即可达到快速、高效地对图像进行识别的目的 .由于卷积神经网络结构众多,且同一数据在不同网络表现也不尽相同,为了选择适用于分析沉船声呐数据的网络,使用Python搭建LeNet 5、AlexNet、VGG、GoogLeNet、ResNet系列(Res...     

BP网络学习参数模糊自适应算法的实现

前馈神经网络BP算法的改进方案中,对网络训练(学习)过程中学习率和惯性系数进行模糊自适应调节,以提高收敛速度,是一项很有效的措施。文中具体分析了如何根据设计者的先验知识确定模糊规则和隶属函数,并以三比特异或函数(或称奇偶分类)的实现为例,验证了这种算法的改进、加速了BP网络的学习过程。