首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到20条相似文献,搜索用时 109 毫秒
1.
深度学习在物理、工程等领域的应用成为人工智能研究的新热点,本文在平抛运动学习中引入深度学习框架,介绍了物理信息神经网络(PINN)的基本理念及使用方法,为进一步探究人工智能在物理学习中的应用提供了参考。  相似文献   

2.
针对模拟电路易发生故障且不易诊断的问题,提出了一种基于深度学习的模拟电路故障诊断算法。该算法首先将采样的原始数据制作成语音形式,然后通过时频域变化转化为语谱图,最后再将其送入VGG16模型中进行训练与测试。实验结果表明,该算法用于模拟电路故障诊断时能够识别的故障种类达到9种,同时准确度达到了100%,具有很强的电路故障诊断能力。  相似文献   

3.
深度学习目前已广泛应用到各个领域,目标检测是计算机视觉领域中的基础问题。针对传统目标检测算法存在的效率低、鲁棒性差等问题,基于深度学习的目标检测算法很好地提高了目标检测效率,成为主流趋势。论文对一些典型的基于深度学习的目标检测算法进行了综述,主要分为基于区域思想和基于回归思想两方面,对算法结构进行了分析和对比,最后对基于深度学习的目标检测算法的发展进行了展望。  相似文献   

4.
5.
目标检测是高级视觉研究领域的重要前提,是计算机视觉研究的核心问题.深度学习拥有强大的自学习能力,将其运用至目标检测领域能够在一定程度上弥补了传统检测方法的不足.首先介绍了传统目标检测方法面临的困境;然后对两阶段深度学习算法和单阶段深度学习算法分别进行介绍;最后对基于深度学习的目标检测算法的发展进行总结,并对未来前景进行...  相似文献   

6.
针对现有疲劳驾驶检测算法实用性差或准确率低的问题,本文提出了一种基于深度学习的疲劳驾驶检测算法.首先,使用HOG (Histogram of Oriented Gradient)特征算子检测人脸的存在;其次,利用特征点模型实现人脸的对齐,同时实现眼睛、嘴巴区域的分割;最后通过深度卷积神经网络提取驾驶员的眼部疲劳特征,并融合驾驶员嘴部的疲劳特征进行疲劳预警.大量的实验表明,该方法在疲劳驾驶检测的准确率、实时性等方面都取得明显的性能提升.  相似文献   

7.
基于深度学习的鱼类分类算法研究   总被引:1,自引:0,他引:1  
回顾近年来国内外对鱼类分类的研究进展,指出传统方法存在的缺陷。深度学习是目前图像分类的主流方法。研究基于卷积神经网络CNN(Convolutional Neural Network)的鱼类分类模型,并以该模型为基础,进一步提出利用迁移学习,以预训练网络的特征结合SVM算法(Pre CNN+SVM)的混合分类模型。实验以Fish4-Knowledge(F4 K)作为数据集,使用Tensor Flow训练网络模型。实验结果表明,利用Pre CNN+SVM算法,取得了98.6%的准确率,较传统方法有显著提高。对于小规模数据集,有效解决了需要人工提取特征的不可迁移性。  相似文献   

8.
图像的自动标注是图像检索领域一项基础而又富有挑战性的任务。深度学习算法自提出以来在图像和文本识别领域取得了巨大的成功,是一种解决\"语义鸿沟\"问题的有效方法。图像标注问题可以分解为基于图像与标签相关关系的基本图像标注和基于标注词汇共生关系的标注改善两个过程。文中将基本图像标注问题视为一个多标记学习问题,图像的标签先验知识作为深度神经网络的监督信息。在得到基本标注词汇的基础上,利用原始图像标签词汇的依赖关系与先验分布改善了图像的标注结果。最后将所提出的改进的深度学习模型应用于Corel和ESP图像数据集,验证了该模型框架及所提出的解决方案的有效性。  相似文献   

9.
深度学习理论是现代计算机机器学习中的非常重要的组成部分.计算机图形描述算法的实现解决了生成算法难的问题.研究了深度学习的理论:卷积神经网络和递归神经网络.结果表明:两种神经网络系统都能很好地处理图像的识别问题,对于计算机图形描述算法有着很好地辅助作用.同时,设计了一种基于卷积神经网络的计算机图形识别算法.实验结果表明:...  相似文献   

10.
随着深度学习的不断发展与图像数据的爆炸式增长,如何使用深度学习来获得更高压缩比和更高质量的图像逐渐成为热点研究问题之一.通过对近几年相关文献的分析与整理,将基于深度学习的图像压缩方法按照卷积神经网络、循环神经网络、生成对抗网络进行总结与分析,对不同种方法分别列举了具有代表性的实例,并对基于深度学习的图像压缩算法的常用训...  相似文献   

11.
《国际计算机数学杂志》2012,89(9):2014-2023
In this article, He's energy balance method (EBM) is applied to solve Van der Pol damped nonlinear oscillators. Three examples of Van der Pol oscillators are presented and solved by this method to illustrate the effectiveness and convenience of the EBM. In this method, only one iteration leads to high accuracy of the solution. Comparisons are made between EBM results and exact solutions of the problems. The results show that the method can be easily extended to other nonlinear systems and can therefore be found widely applicable in engineering and other science.  相似文献   

12.
本文研究了不确定Van der Pol混沌系统的同步问题,并进行了基于规则的模糊逻辑控制器(FLC)的控制。首先寻找主从Van der Pol混沌系统满足Lyapunov稳定性理论的条件,在此基础上建立模糊规则,设计模糊控制器,实现不确定混沌系统的同步。通过不确定VanderPol混沌系统的两组仿真结果,验证了模糊同步控制方法具有很好的鲁棒性。最后为了进一步验证该方法的有效性,本文在相同条件下,利用反馈控制的方法实现不确定主从VanderPol混沌系统的同步,然后再将此方法的仿真结果与本文的模糊同步控制方法的仿真结果在稳态误差及同步所需时间这两个方面进行对比分析。分析结果验证了本文同步方法的可行性及有效性。  相似文献   

13.
研究多自由度Van der Pol型非线性振动系统振幅增大的控制,设计反馈控制器,用数值方法对控制系统的幅值进行了计算,绘制了在不同控制参数下,系统响应的时间历程曲线和极限环.研究表明通过调整控制参数,能够增大极限环的幅值,有工程应用价值,对高维系统的分岔控制研究有一定的理论意义.  相似文献   

14.
    
Abstract: The concept of linked oscillators in biological control systems has long been established. Frequency entrainment is a predominant explanation behind many biological rhythms. In this paper a preliminary examination of electroencephalographic entrainment is made to survey the possibility and methods of achieving signal entrainment at the highest level of neurological organization and function. A model of the thalamocortical system is employed to generate simulated electroencephalographic signals and is tested in various configurations in the search for entrainment under very simple conditions. Additionally, an analysis of the coupled Van der Pol model of the circadian rhythm controller is performed to identify the possibility of affecting that system with a drastically different coupling input signal. We were able to conclude that overall signal shape can have a significant impact on the entrainment characteristics of the system. Due to the nature of the underlying mathematical structure of the model, by examining the circadian rhythm controller, we found that it is unsuitable for entrainment to an incident entraining signal of much higher frequency.  相似文献   

15.
深度神经网络模型压缩综述   总被引:1,自引:0,他引:1  
近年来;随着深度学习的飞速发展;深度神经网络受到了越来越多的关注;在许多应用领域取得了显著效果。通常;在较高的计算量下;深度神经网络的学习能力随着网络层深度的增加而不断提高;因此深度神经网络在大型数据集上的表现非常卓越。然而;由于其计算量大、存储成本高、模型复杂等特性;使得深度学习无法有效地应用于轻量级移动便携设备。因此;压缩、优化深度学习模型成为目前研究的热点。当前主要的模型压缩方法有模型裁剪、轻量级网络设计、知识蒸馏、量化、体系结构搜索等。对以上方法的性能、优缺点和最新研究成果进行了分析总结;并对未来研究方向进行了展望。  相似文献   

16.
    
This paper proposes an asymptotic rejection algorithm for rejecting exotic disturbances in nonlinear systems. The disturbances, which are produced by nonlinear exosystems, are nonharmonic and periodic. A new internal model is proposed to deal with the disturbances. Further, an adaptive output feedback controller is designed to ensure that the system's state variables can asymptotically converge to zero, and the disturbances can be completely rejected. The proposed algorithm can be used in many applications, e.g. active vibration control, and the avoidance of nonharmonic distortion in nonlinear circuits. An example is performed to demonstrate that the proposed algorithm can completely reject the nonharmonic periodic disturbances produced by nonlinear exosystems. Copyright © 2010 John Wiley and Sons Asia Pte Ltd and Chinese Automatic Control Society  相似文献   

17.
The one-dimensional linear wave equation with a van der Pol nonlinear boundary condition is one of the simplest models that may cause isotropic or nonisotropic chaotic vibrations. It characterizes the nonisotropic chaotic vibration by means of the total variation theory. Some results are derived on the exponential growth of total variation of the snapshots on the spatial interval in the long-time horizon when the map and the initial condition satisfy some conditions.  相似文献   

18.
    
Fingerprint classification is one of the most common approaches to accelerate the identification in large databases of fingerprints. Fingerprints are grouped into disjoint classes, so that an input fingerprint is compared only with those belonging to the predicted class, reducing the penetration rate of the search. The classification procedure usually starts by the extraction of features from the fingerprint image, frequently based on visual characteristics. In this work, we propose an approach to fingerprint classification using convolutional neural networks, which avoid the necessity of an explicit feature extraction process by incorporating the image processing within the training of the classifier. Furthermore, such an approach is able to predict a class even for low‐quality fingerprints that are rejected by commonly used algorithms, such as FingerCode. The study gives special importance to the robustness of the classification for different impressions of the same fingerprint, aiming to minimize the penetration in the database. In our experiments, convolutional neural networks yielded better accuracy and penetration rate than state‐of‐the‐art classifiers based on explicit feature extraction. The tested networks also improved on the runtime, as a result of the joint optimization of both feature extraction and classification.  相似文献   

19.
环面上van der Pol方程混沌解的可视化计算   总被引:2,自引:0,他引:2  
研究了环面上非线性van der Pol方程的图形建模及可视化计算的问题,在图形环境下,对van der Pol环面方程从建模,实验到结果分析的全过程进行了可视化建模和计算,并建立了一个对系统运动轨迹进行全面自动分析试验的可视化仿真框架,该方法不但可以不用传统程序代码对模型及算法编程,而且可方便地对系统进行多参数自动迭代试验及智能分析。  相似文献   

20.
Using a combination of analytical and numerical methods, the global bifurcations and chaotic dynamics of two non-linearly coupled parametrically excited van der Pol oscillators are investigated in detail. With the aid of the method of multiple scales, the slow flow equations are obtained. Based on the slow flow equations, normal form theory and the techniques of choosing complementary space are applied to find the explicit expressions of the simpler normal form associated with a double zero and a pair of pure imaginary eigenvalues. By the simpler normal form, using the global perturbation method developed by Kovacic and Wiggins, the analysis of global bifurcation and chaotic dynamics of two non-linearly coupled parametrically excited van der Pol oscillators is given. The results indicate that there exists a Silnikov type single-pulse homoclinic orbit for this class of system which implies the chaotic motions can occur. Numerical simulations are also given and verify the analytical predictions.  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号