高阶CMAC神经网络及其在水下机器人运动控制中的应用

对高阶ＣＭＡＣ神经网络的结构和工作原理进行了研究,提出了中间层作用函数地址的计算方法,给出了计算高阶基函数的不同方法,利用高阶ＭＣＡＣ神经网络对水下机器人模糊深度控制器进行了学习,仿真结果显示了不同高阶基函数ＭＣＡＣ网络的不同建模能力,证明了中间层作用函数地址的计算方法正确。     

基于神经网络的无人无缆水下机器人运动建模与控制技术研究

对自主型水下机器人（ＡＵＶ）神经网络运动模型的结构进行了理论分析和探讨,提出了非完全回归型神经网络、增加积分层的输出层结构及相应的分步式学习方法。对ＡＵＶ运动过程中目标运动路径和目标运动速度的同时跟踪控制进行了系统研究。提出了由主控网络和伴随网络构成的神经网络控制器结构,给出了通过计算机模拟来生成教师样本的方法,提出了预测控制的思想。计算机仿真及水下机器人“Ｔｗｉｎ－Ｂｕｒｇｅｒ”的水池实验结果验证了本文所提出的建模方法和跟踪控制方法的有效性和可行性。     

水下机器人非完全回归型神经网络运动模型的研究

研究了基于神经网络建立水下机器人（ＡＵＶ）运动模型的方法,提出以非完全回归形式来表现输入输出变量的时序列影响和输出层向输入层回归的封闭式结构以实现运动模型的模拟机能,并给出了相应的训练和学习方法。此外,又提出了对输出层增加积分环节及神经网络输入输出变量进行正规化处理的方法。计算机的仿真及实际水下机器人“ＰＷ４５”的建模结果,表明本文所提出的建模方法的有效性和可行性。     

基于动态神经网络的AUV航向自适应控制

海洋环境的复杂性以及自身模型的不确定性,给自主潜航器(AUV)航向控制带来很大困难.针对AUV的特点和控制方面所存在的问题,采用了带衡量因子的动态BP神经网络控制器控制AUV的航向.理论分析和仿真结果表明,与传统的PID控制器相比,在扰动存在的情况下,神经网络控制器具有更好的自适应性和鲁棒性.     

局部回归神经网络在水下机器人运动控制中的应用

探讨了水下机器人神经网络运动模型的结构,提出了带有局部回归结构的水下机器人神经网络控制器结构及预测控制的实现方法,给出了该神经网络的教师样本生成方法及学习方法,计算机仿真与水池实验结果验证了本文提出方法的有效性和可行性。     

人工神经网络控制器在水下无人运载器导航中的应用

本文以水下无人运载器（AUV）绕平面圆周航行的速度、位置控制为例，推广到任何轨迹的控制问题。利用神经网络学习AUV运行的内在规律，预测未来一步的运行情况，并用改良的PID方式前馈与后馈相结合控制其执行机构。系统学习、预测及PID各增益量利用GESA（GuidedEvolutionarySimulatedAnnealing）全局优化方法求得。本方法具有自适应性、强非线性及前馈控制等特点，优于其它一般的控制器。     

基于强化学习的自主式水下潜器障碍规避技术（英文）

Obstacle avoidance becomes a very challenging task for an autonomous underwater vehicle(AUV) in an unknown underwater environment during exploration process. Successful control in such case may be achieved using the model-based classical control techniques like PID and MPC but it required an accurate mathematical model of AUV and may fail due to parametric uncertainties, disturbance, or plant model mismatch. On the other hand, model-free reinforcement learning(RL) algorithm can be designed using actual behavior of AUV plant in an unknown environment and the learned control may not get affected by model uncertainties like a classical control approach. Unlike model-based control model-free RL based controller does not require to manually tune controller with the changing environment. A standard RL based one-step Q-learning based control can be utilized for obstacle avoidance but it has tendency to explore all possible actions at given state which may increase number of collision.Hence a modified Q-learning based control approach is proposed to deal with these problems in unknown environment.Furthermore, function approximation is utilized using neural network(NN) to overcome the continuous states and large statespace problems which arise in RL-based controller design. The proposed modified Q-learning algorithm is validated using MATLAB simulations by comparing it with standard Q-learning algorithm for single obstacle avoidance. Also, the same algorithm is utilized to deal with multiple obstacle avoidance problems.     

AUV hull shape design based on desired pressure distribution

The role of autonomous underwater vehicles (AUVs) is more important in the quest to reach the deep seas today than ever before. The hull shape of the AUV can differ depending on the special mission considered for the vehicle. Therefore, different types of algorithms for the body shape design of these kinds of vehicles are being developed every day. In the current work, a new procedure has been proposed for designing the body shape of an AUV. Using this method which is based on a desired pressure distribution, it is possible to obtain the desired hull shape design. Artificial neural network algorithm has been used for this purpose. Preliminary data for training and testing of the network have been obtained from CFD simulation of the flow around the body of Hydrolab500 AUV. In this regard, pressure distribution has been evaluated around each body by changing the nose and tail profile of AUV. The results obtained from this research indicate that a body correlated to the desired pressure can be designed properly.     

Expert S-surface control for autonomous underwater vehicles

S-surface control has proven to be an effective means for motion control of underwater autonomous vehicles （AUV）. However there are still problems maintaining steady precision of course due to the constant need to adjust parameters, especially where there are disturbing currents. Thus an intelligent integral was introduced to improve precision. An expert S-surface control was developed to tune the parameters on-line, based on the expert system, it provides S-surface control according to practical experience and control knowledge. To prevent control output over-compensation, a fuzzy neural network was included to adjust the production rules to the knowledge base. Experiments were conducted on an AUV simulation platform, and the results show that the expert S-surface controller performs better than an S-surface controller in environments with currents, producing good steady precision of course in a robust way.     

基于模糊神经网络的水下机器人局部路径规划方法

该文探讨了基于模糊神经网络理论的实时局部路径规划问题,并提出了能实现模糊控制规则的基于强化学习的自学习和自调整的规划算法,设计了水下机器人实时运动规划器结构以及规划器操作过程和相应的算法。仿真实验结果验证了本文所提方法的有效性和可行性。     

模糊解耦在UUV回收运动控制中的应用

在回收过程中,UUV (Unmanned Underwater Vehicle) 对水平面运动的控制精度有很高的要求。以BSA-UUV为平台,构建水平面操纵非线性方程,在此模型的基础上,分析了多阶段回收过程中的主要耦合变量和耦合原因。针对水平面运动中航向控制与横向运动之间的强耦合问题,基于模糊理论和解耦理论设计一种解耦补偿器,由模糊补偿器的输入输出隶属度函数,根据模糊补偿规则,经过模糊推理合成运算和清晰化运算,得出解耦补偿量。仿真结果显示加入模糊解耦控制器以后,有效降低了系统的超调量,提高了控制精度,表明模糊解耦控制方法在UUV回收运动控制中有很高的应用价值。     

基于改进A*GBNN算法的AUV路径规划研究

针对现有的离散生物启发神经网络(Glasius bioinspired neural networks, GBNN)算法在未知环境下,存在的路径规划时间长、易陷入局部最优等问题,提出一种结合A*与GBNN模型的改进算法。在GBNN活性值栅格网络中,算法将各栅格的活性值作为A*的代价函数进行运算并使用跳点搜索规则优化,实现未知环境下的实时路径规划。仿真实验结果表明,该算法有效改善了自主水下航行器在未知环境下的寻路效率,可以满足自主水下航行器实时路径规划需求。     

基于卷积神经网络的AUV水下识别算法设计与实现

自主式水下机器人(AUV)是应用于复杂海洋环境中的高智能化无人装备,其需要具备良好的环境感知能力进行自主导航,包括水下目标识别能力。随着人工智能的高速发展,卷积神经网络作为图像处理领域的深度学习架构,在图像特征提取和图像识别上有着强大的性能和卓越的优势。本文利用卷积神经网络,实现了自主式水下机器人水下目标的自主识别。同时,通过采用三段式全连接方式和增加卷积层深度的方式对卷积神经网络进行进一步改进,提高了卷积神经网络的训练速度、准确率和泛化能力。     

水下智能潜器分布式控制系统设计与实现

分布式控制系统是水下管能潜器的重要部分．本文采用分层控制结构，即执行级、协调级和组织级，实现了潜器的控制系统。文中首先介绍了控制系统的组成，然后分别介绍了执行级采用神经网络实现自适应控制的方法，协调级中实现全局和局部路径规划的方法，以及组织级中基于行为分解的任务规划方法。文中给出了系统的仿真和实验结果。     

一种基于PSO优化HWFCM的快速水下图像分割算法

The S/N of an underwater image is low and has a fuzzy edge.If using traditional methods to process it directly,the result is not satisfying.Though the traditional fuzzy C-means algorithm can sometimes divide the image into object and background,its time-consuming computation is often an obstacle.The mission of the vision system of an autonomous underwater vehicle (AUV) is to rapidly and exactly deal with the information about the object in a complex environment for the AUV to use the obtained result to execute the next task.So,by using the statistical characteristics of the gray image histogram,a fast and effective fuzzy C-means underwater image segmentation algorithm was presented.With the weighted histogram modifying the fuzzy membership,the above algorithm can not only cut down on a large amount of data processing and storage during the computation process compared with the traditional algorithm,so as to speed up the efficiency of the segmentation,but also improve the quality of underwater image segmentation.Finally,particle swarm optimization (PSO) described by the sine function was introduced to the algorithm mentioned above.It made up for the shortcomings that the FCM algorithm can not get the global optimal solution.Thus,on the one hand,it considers the global impact and achieves the local optimal solution,and on the other hand,further greatly increases the computing speed.Experimental results indicate that the novel algorithm can reach a better segmentation quality and the processing time of each image is reduced.They enhance efficiency and satisfy the requirements of a highly effective,real-time AUV.     

模糊神经网络在船舶火灾报警系统中的应用

提出了一种利用神经网络和模糊系统理论设计船舶火灾报警系统的方法。该系统 模糊推理能力哟、神经网络学习能力强的特点,将模糊系统和神经网络结合起来,实现模糊系统设计的自动调整。利用感温式知因式传感器对火灾发生时温度和烟雾两个主要参量进行智能监测委 解决了报警灵敏度和误报率之间的矛盾。     

基于模糊RBF神经网络的辐射源识别

针对模糊识别系统的不足,为了提高辐射源识别系统的识别正确率,构建了基于模糊RBF神经网络的辐射源识别系统,提出了一种等价型模糊RBF神经网络的结构和学习算法,采用五层神经网络结构来实现模糊系统的模糊化和规则推理,神经网络的所有节点和参数对应了模糊系统的隶属函数和推理过程.在仿真实验中,分别采用模糊识别系统、并联型模糊RBF神经网络、结构等价型模糊RBF神经网络进行辐射源识别,给出了三种算法在相同噪声环境下的仿真结果,表明等价型模糊RBF效神经网络有较高的正确识别率,具有更强的抗干扰能力,但运算量相对较大.     

带吊舱水下机器人运动建模与仿真(英文)

To provide a simulation system platform for designing and debugging a small autonomous underwater vehicle’s (AUV) motion controller, a six-degree of freedom (6-DOF) dynamic model for AUV controlled by thruster and fins with appendages is examined. Based on the dynamic model, a simulation system for the AUV’s motion is established. The different kinds of typical motions are simulated to analyze the motion performance and the maneuverability of the AUV. In order to evaluate the influences of appendages on the motion performance of the AUV, simulations of the AUV with and without appendages are performed and compared. The results demonstrate the AUV has good maneuverability with and without appendages.     

基于IGA-BP算法的船舶航向智能自适应控制系统设计

在深入研究基于BP学习算法的前向神经网络以及模糊神经网络控制器的基础上,针对模糊神经网络控制器难以设计以及传统BP学习算法易于陷入局部收敛的不足,结合免疫遗传算法的全局收敛特性以及BP学习算法局部收敛的快速性,提出了一种基于混合计算智能方法的IGA-BP算法的神经网络参数的优化设计方法.将设计的控制器用于建立船舶航向控制系统模型,仿真结果表明,在船舶无干扰和存在随机干扰的情况下,基于IGA-BP算法设计的船舶航向控制系统均能使船舶转向控制无超调,跟踪快,比BP学习算法的控制效果更理想.     

基于ASM和FNN的人脸识别

将主动形状模型ASM(Active Shape Model)与模糊神经网络FNN(Fuzzy Neural Networks)相结合并用于人脸识别。针对经典ASM对初始化位置敏感、收缩速度慢的现象和传统模糊神经网络难以解决高维模式分类的问题进行了分析和改进。ORL人脸库上的试验表明了该算法的有效性。