基于深度卷积神经网络的剪切波分裂质量检测     

张彭达  戴志阳  查显杰 《中国地震》2020,36(3):539-549

剪切波分裂是分析地震各向异性的一种重要手段,常规方法是利用网格搜索获取分裂参数,再通过不同方法的测量结果对比测量结果进行质量检测,这一过程会耗费大量计算时间。本文针对这一问题提出了一种利用深度卷积神经网络对剪切波分裂进行质量检测的新方法,对使用了Resnet残差结构的深度神经网络进行训练,直接对二分量剪切波波形数据的质量进行分类。整个过程为:神经网络通过卷积层提取波形特征,计算损失函数后反向传播训练模型参数,完成迭代训练后的模型对输入波形数据正向计算自动输出类型。本文利用川西台站接收到的实际数据以及随机生成的合成数据分别对该网络进行训练,均可以获得准确的分类结果。相比于通过多种剪切波分裂方法对比测量结果的质量检测方法,基于神经网络的方法可以省略网格搜索的计算过程直接判断质量类型,在运算速度上的优势明显,并可继续通过训练提高模型的精度,为提升剪切波分裂方法在数据处理过程中的操作效率提供帮助。  相似文献   

基于遗传算法优化BP神经网络的玉米遥感估产方法     

于海洋  陈圣波  杨北萍  安秦 《世界地质》2020,39(1):208-214

以内蒙古自治区开鲁县玉米作物为研究对象,将生育期内玉米遥感影像所提取的多种植被指数和实地采样点的测产数据作为训练值,利用BP(back propagation)神经网络和遗传算法优化BP(GA-BP)神经网络估产模型,得出网络预测的玉米产量数值。通过决定系数R 2和均方根误差RMSE,比较实测产量与预测产量之间的精度,BP神经网络模型R^2为0.8452,RMSE(%)为28.37;遗传算法优化BP神经网络模型R^2为0.9850,RMSE(%)为6.70,表明遗传算法优化BP神经网络估产模型具有一定可行性和可信度。  相似文献   

Simulation of urban expansion via integrating artificial neural network with Markov chain – cellular automata     

Tingting Xu  Jay Gao  Giovanni Coco 《International journal of geographical information science》2019,33(10):1960-1983

Accurate simulations and predictions of urban expansion are critical to manage urbanization and explicitly address the spatiotemporal trends and distributions of urban expansion. Cellular Automata integrated Markov Chain (CA-MC) is one of the most frequently used models for this purpose. However, the urban suitability index (USI) map produced from the conventional CA-MC is either affected by human bias or cannot accurately reflect the possible nonlinear relations between driving factors and urban expansion. To overcome these limitations, a machine learning model (Artificial Neural Network, ANN) was integrated with CA-MC instead of the commonly used Analytical Hierarchy Process (AHP) and Logistic Regression (LR) CA-MC models. The ANN was optimized to create the USI map and then integrated with CA-MC to spatially allocate urban expansion cells. The validated results of kappa and fuzzy kappa simulation indicate that ANN-CA-MC outperformed other variously coupled CA-MC modelling approaches. Based on the ANN-CA-MC model, the urban area in South Auckland is predicted to expand to 1340.55 ha in 2026 at the expense of non-urban areas, mostly grassland and open-bare land. Most of the future expansion will take place within the planned new urban growth zone.  相似文献   

经济地理学视角下创新网络与区域增长研究述评     

曹贤忠  曾刚 《热带地理》2019,39(3):472-478

创新是引领经济发展的第一动力，创新与区域增长之间的关系成为经济地理学者关注的重点领域。文章通过梳理近年来有关创新网络测度、创新网络与区域增长关系、创新网络作用于区域增长方式等方面的文献发现：网络资本可以弥补社会资本在解释企业创新结网经济价值方面的不足，区域增长呈现出网络化特征已成为学界共识，知识流与邻近性能较好地解释创新网络与区域增长的关系机理。然而，当前研究对社会资本如何促进区域增长，网络资本与区域增长关系模型如何构建，不同类型的邻近性与知识对区域增长的影响有何差异等问题尚不明确，建议重视网络资本对区域增长的作用并实证检验二者的关联，同时还应重视创新网络中企业家精神、创新个体心理行为特征等因素对区域增长的影响。  相似文献   

时间尺度重构EEMD-GRNN改进模型预测PM2.5的研究     

符海月  张祎婷 《地球信息科学学报》2019,21(7):1132-1142

合理构建PM_2.5浓度预测模型是科学、准确地预测PM_2.5浓度变化的关键。传统PM_2.5预测EEMD-GRNN模型具有较好的预测精度,但是存在过于关注研究数据本身而忽略其物理意义的不足。本研究基于南京市2014-2017年PM_2.5浓度时间序列数据,分析PM_2.5浓度多尺度变化特征及其对气象因子和大气污染因子的尺度响应,基于时间尺度重构进行EEMD-GRNN模型的改进与实证研究。南京市样本数据PM_2.5浓度变化表现为明显的天际尺度和月际尺度,从重构尺度（天际、月际）构建GRNN模型更具有现实意义;同时,PM_2.5对PM₁₀、NO₂、O₃、RH、MinT等因子存在多尺度响应效应,以其作为GRNN模型中的输入变量更具有时间序列上的解释意义。改进后的EEMD-GRNN模型具有更高的PM_2.5浓度预测精度,MAE、MAPE、RMSE和R²分别为6.17、18.41%、8.32和0.95,而传统EEMD-GRNN模型的模型有效性检验结果分别为8.37、27.56%、11.56、0.91。对于高浓度天（PM_2.5浓度大于100 μg/m³）的预测,改进模型更是全面优于传统EEMD-GRNN模型,MAPE为12.02%,相较于传统模型提高了9.03%。  相似文献   

Evaluating techniques for mapping island vegetation from unmanned aerial vehicle (UAV) images: Pixel classification,visual interpretation and machine learning approaches     

《International Journal of Applied Earth Observation and Geoinformation》2020

We evaluate three approaches to mapping vegetation using images collected by an unmanned aerial vehicle (UAV) to monitor rehabilitation activities in the Five Islands Nature Reserve, Wollongong (Australia). Between April 2017 and July 2018, four aerial surveys of Big Island were undertaken to map changes to island vegetation following helicopter herbicide sprays to eradicate weeds, including the creeper Coastal Morning Glory (Ipomoea cairica) and Kikuyu Grass (Cenchrus clandestinus). The spraying was followed by a large scale planting campaign to introduce native plants, such as tussocks of Spiny-headed Mat-rush (Lomandra longifolia). Three approaches to mapping vegetation were evaluated, including: (i) a pixel-based image classification algorithm applied to the composite spectral wavebands of the images collected, (ii) manual digitisation of vegetation directly from images based on visual interpretation, and (iii) the application of a machine learning algorithm, LeNet, based on a deep learning convolutional neural network (CNN) for detecting planted Lomandra tussocks. The uncertainty of each approach was assessed via comparison against an independently collected field dataset. Each of the vegetation mapping approaches had a comparable accuracy; for a selected weed management and planting area, the overall accuracies were 82 %, 91 % and 85 % respectively for the pixel based image classification, the visual interpretation / digitisation and the CNN machine learning algorithm. At the scale of the whole island, statistically significant differences in the performance of the three approaches to mapping Lomandra plants were detected via ANOVA. The manual digitisation took a longer time to perform than others. The three approaches resulted in markedly different vegetation maps characterised by different digital data formats, which offered fundamentally different types of information on vegetation character. We draw attention to the need to consider how different digital map products will be used for vegetation management (e.g. monitoring the health individual species or a broader profile of the community). Where individual plants are to be monitored over time, a feature-based approach that represents plants as vector points is appropriate. The CNN approach emerged as a promising technique in this regard as it leveraged spatial information from the UAV images within the architecture of the learning framework by enforcing a local connectivity pattern between neurons of adjacent layers to incorporate the spatial relationships between features that comprised the shape of the Lomandra tussocks detected.  相似文献   

A-FPN算法及其在遥感图像船舶检测中的应用     

于野  艾华  贺小军  于树海  钟兴  朱瑞飞 《遥感学报》2020,24(2):107-115

光学遥感图像船舶检测主要面临两个挑战:光学遥感图像背景复杂,船舶检测易受海浪、云雾及陆地建筑等多方面干扰;遥感图像分辨率低,船舶目标小,对于其分类与定位带来很大困难;针对上述问题,在FPN的基础上,提出一种融入显著性特征的卷积神经网络模型A-FPN (Attention-Based Feature Pyramid Networks)。首先,利用卷积提取图像特征金字塔;然后,利用顶层金字塔逐级构建显著特征层,抑制背景信息,通过金字塔顶层的细粒度特征提高浅层特征的表达能力,构建自上而下的多级显著特征映射结构;最后利用Softmax分类器进行多层级船舶检测。A-FPN模型利用显著性机制引导不同感受下的特征进行融合,提高了模型的分辨能力,对遥感图像处理领域具有重要应用价值。实验阶段,利用公开的遥感目标检测数据集NWPU VHR-10中的船舶样本进行测试,准确率为92.8%,表明A-FPN模型适用于遥感图像船舶检测。  相似文献   

不同结构深度神经网络的时间域航空电磁数据成像性能分析     

李金峰  刘云鹤 《世界地质》2020,39(1):159-166

时间域航空电磁系统采样密集,数据量大,所以在该领域较为实用的数据处理方法主要为一维反演和电阻率成像法。笔者从成像问题出发,建立了庞大的数据模型训练集,研究并分析了不同结构的神经网络的成像精度。通过对比分析测试结果,获得了在一定条件下适用于航空电磁成像的最优网络模型结构,包含其神经元个数和层数等信息。本文采用早停法训练神经网络,压制数据中噪声对成像结果的影响。  相似文献   

基于小波变换与神经网络的石羊河流域夏季地温预测模型研究   总被引：1，自引：1，他引：0  

贾东于  李开明  聂晓英  袁春霞  李清峰  高福元 《冰川冻土》2020,42(2):412-422

地温变化在气候反馈效应中起着重要作用， 理解地温及其与影响因素之间的时空关系对预测全球温度变化至关重要。利用1998 - 2017年石羊河流域的逐日常规气象观测资料， 采用小波分析结合BP（Back Propagation）神经网络构建了石羊河流域夏季地温预报模型， 结果表明： 日平均地温预测效果在不同站点均为最佳， 其中预测值和观测值的相关系数均大于0.87， 3 ℃以内的预测概率均大于84%。其中， 民勤地区地温预测效果最好， 预测值和观测值的相关系数达到0.91， 3 ℃以内的预测概率达到86%。日最高地温的预测值与观测值的相关系数高于0.8， 但误差平方和、 标准差较大。永昌地区日最高地温的模拟效果最好， 3 ℃以内的预测概率达到83%。日最低地温的预测与观测值的平均相关系数高于0.66， 3 ℃以内的预报概率高于83%， 但预测值略低。其中， 武威地区日最低地温的预测效果最好， 预测值与观测值的相关系数为0.72， 3 ℃以内的预测概率达到94%。研究成果可为有效弥补干旱、 半干旱区地温观测资料缺失和探讨其与局地气候的关系提供一些参考。  相似文献   

Analysis of resistance factors for LRFD of soil nail pullout limit state using default FHWA load and resistance models     

Hui Hu 《Marine Georesources & Geotechnology》2020,38(3):332-348

Abstract

Resistance factors for load and resistance factor design (LRFD) of pullout limit state of both permanent and temporary soil nails are calibrated against a wide design space using the current Federal Highway Administration (FHWA) nail load and resistance models. The calculated resistance factors were shown to scatter broadly among design scenarios that differ in wall face batter, soil friction angle, nail ultimate bond strength, and surcharge live load. An important lesson learned from the analysis results is that the current practice of using a single resistance factor for LRFD of nail pullout limit state could not result in uniform reliabilities across different design scenarios. Simple artificial neural network (ANN) models were developed for computation of resistance factors. Design examples demonstrated the ability of the ANN models in providing resistance factors that yield satisfactory and consistent reliabilities in different nail pullout designs.  相似文献