基于遗传算法优化的小波神经网络在地下水位预测中的应用

地下水位是衡量生态环境优劣和地下水资源的一个重要指标.地下水位下降,将引发地面沉降、地面塌陷和降落漏斗等.因此,地下水位预测对保护地质生态环境和实现地下水资源严格管理至关重要.由于BP算法存在极易收敛于局部极小点与过拟合等缺点,导致网络泛化能力不足,本文在构建小波神经网络基础上并引入遗传算法加以优化,以解决上述不足,并与BP和WNN对比预测了天津市深层承压水水位.预测结果表明,GA-WNN模型拟合精度较高,模型的预测能力有较大幅度提高.     

Carbon and Oxygen Stable Isotope Values and Microfossils at 41.4–4.5 ka BP in Tai Co, Tibet, China, and Their Paleoclimatic Significance

Two sedimentary sections were measured at Tai Co (Co means lake in Tibetan) in western Tibet, China. The two sections are almost all composed of clay carbonate beds except in their lower parts where there are carbonized plants at >10 cm depths and dark-colored carbonate clay and clay at 50–70 cm depths, yielding abundant gastropods, ostracods, and charophytes. The carbon and oxygen stable isotope values of carbonate, ostracods, and charophytes and ecological features of microfossil communities indicate the following climatic change in the area from 41.4 to 4.5 ka BP: at 41.4–26.2 ka BP, the climate was relatively wet; at 26.2–25.5 ka BP, it was slightly warm-dry; at 25.5–22.5 ka BP, it was warm-wet; at 22.5–21.0 ka BP, it was slightly cold-wet; at 20.5–17.5 ka BP, it became cold abruptly and slightly wet, implying the climate of the last glacial maximum; at 17.5–16.0 ka BP, it was slightly cold-dry; at 16.0–11.8 ka BP, it was slightly warm-wet; at 11.8–10.4 ka BP, it was relatively cold-dry, roughly equivalent to the climate of the Younger Dryas, and at 10.4 ka BP, the temperatures began to rise again; at 10.4–9.4 ka BP, it was slightly warm-wet; at 9.4–8.5 ka BP, there occurred short warm-wet oscillations; at 8.5–7.9 ka BP, it was slightly dry-cold, representing a strong temperature-lowering and drying event in the postglacial stage; at 7.8–6.3 ka BP, it was slightly warm-wet; at 6.3–4.5 ka BP, the climate tended to be cold-dry. 4.5 ka BP recorded the maximum aridity since the late part of the late Pleistocene.     

Optimal design of structures for earthquake loads by a hybrid RBF-BPSO method

The optimal seismic design of structures requires that time history analyses （THA） be carried out repeatedly. This makes the optimal design process inefficient, in particular, if an evolutionary algorithm is used. To reduce the overall time required for structural optimization, two artificial intelligence strategies are employed. In the first strategy, radial basis function （RBF） neural networks are used to predict the time history responses of structures in the optimization flow. In the second strategy, a binary particle swarm optimization （BPSO） is used to find the optimum design. Combining the RBF and BPSO, a hybrid RBF-BPSO optimization method is proposed in this paper, which achieves fast optimization with high computational performance. Two examples are presented and compared to determine the optimal weight of structures under earthquake loadings using both exact and approximate analyses. The numerical results demonstrate the computational advantages and effectiveness of the proposed hybrid RBF-BPSO optimization method for the seismic design of structures.     

灰色模型与智能算法组合模型在变形预测中的应用

变形监测工程是一个复杂的综合系统,各种参数具有很大的不确定性.目前变形的预测分析多采用单一的预测方法,而各种方法都有各自的优缺点和应用范围,有时单一的预测方法对判定工程性质带来了困难.引入了组合预测的思想,在灰色GM(1,1)模型的基础上,构建了灰色+GA+BP神经网络组合模型,探索了时间序列的数据处理和预报问题,通过...     

基于EIGEN-6C4的区域重力异常插值方法

重力异常作为地球重力场的一种基本指标,广泛应用于大地测量、地球物理、地质、地震与海洋领域。为了快速获取高精度的未知点重力异常,本文基于EIGEN-6C4重力场模型解算某110 km×110 km区域内重力异常,构建5′×5′的重力异常格网模型,结合格网节点数据采用6种不同的插值算法拟合未知点的重力异常,最后使用外部检核的方法评价插值算法的精度。实验结果表明,6种插值算法的拟合精度均达到4 mgal以内,自然三次样条法和反距离加权插值法拟合精度最高,得到精度达到2.7 magl的重力异常。     

Ternary forecast of heavy snowfall in the Honam area,Korea

The objective of this study is to improve the statistical modeling for the ternary forecast of heavy snowfall in the Honam area in Korea. The ternary forecast of heavy snowfall consists of one of three values, 0 for less than 50 mm, 1 for an advisory (50–150 mm), and 2 for a warning (more than 150 mm). For our study, the observed daily snow amounts and the numerical model outputs for 45 synoptic factors at 17 stations in the Honam area during 5 years (2001 to 2005) are used as observations and potential pre...     

基于深度学习的天气雷达回波序列外推及效果分析

天气雷达探测资料是进行强对流天气临近预报的主要参考数据。针对传统雷达回波外推方法中存在资料信息利用率不足和外推时效有限的问题,文中利用神经网络进行雷达回波的外推、利用预测神经网络模型进行2 h以内的回波变化预报。回波外推问题的关键是回波时、空序列预测问题,该网络具有解决时间记忆问题的长、短时记忆单元（Long Short-Term Memory,LSTM）和提取空间特征的卷积模块。应用福建、江苏和河南多年的雷达探测资料构造训练和测试数据集。为消除降水的不平衡和提高对强回波的预报准确率,网络采用带权重的损失函数进行训练。对光流法和预测神经网络进行测试集检验以及个例分析,结果表明,在相同外推时效和检验反射率阈值的情况下,预测神经网络的临界成功指数、命中率均高于光流法,虚警率低于光流法。不同类型降水预测神经网络的SSIM值（structural similarity）均高于光流法,且层状云降水的SSIM值比对流云降水的大。因此,预测神经网络对强回波的预报能力高于光流法;在预报时效性上,预测神经网络模型具有一定的优越性;预测神经网络对层状云降水预报的准确率比对流云降水的高。      

塔河油田下奥陶统碳酸盐岩储层测井分类识别研究

塔河油田下奥陶统碳酸盐岩为一套缝洞发育的复杂储层,通过大量已知井的测井曲线与岩心分析数据对照观察、分析,归纳出了五种储层的电性分类识别标准。在此基础上,又利用自然伽玛(GR)、深浅侧向电阻率(Rd,Rs)、声波时差(AC)、密度(DEN)、中子孔隙度(CNL)的响应特征,建立了神经网络识别模型。通过在该区三十多口井的实际应用,取得了较好的效果。     

A-FPN算法及其在遥感图像船舶检测中的应用

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。     

改进的约束总体最小二乘法

针对传统的约束最小二乘模型和总体最小二乘模型的局限性,该文提出了一种改进的约束总体最小二乘法。假设约束总体最小二乘问题中约束方程系数矩阵也存在误差,然后构造函数模型的广义拉格朗日函数,采用最小二乘法迭代求解非线性的法方程,最终获得了改进的约束总体最小二乘法的牛顿-高斯迭代公式和平差模型精度的无偏估计。该算法采用了更接近实际的平差模型,能够获得更加接近真值的估计参数,同时平差模型的精度更加接近模拟数据加入的噪声水平。实验结果表明,本文算法可有效解决对参数进行约束时的数据处理问题。