基于LM-BP神经网络的耕地土壤养分等级划分模型──以皖南山区为例

研究旨在通过BP神经网络方法,构建起LM-BP网络结构(5-M-1)模型,达到对土壤养分等级划分的目的,为合理的土壤养分管理提供可靠依据。采用Levenberg-Marquardt (LM)训练算法,构建3层网络模型：一个输入层、一个隐含层、一个输出层,利用3层网络作为耕地土壤养分等级划分模型。利用土壤养分各级评价标准作为模型的训练样本和测试样本,以此来对BP神经网络进行训练和测试,并对歙县土壤养分进行综合评价。结果表明：LM-BP网络结构对测试样本输出的预测值和实际参考值是一致的。最终通过灰色关联模型和主成分分析方法对歙县土壤养分的综合评价结果与BP神经网络的模拟结果相对比,发现也是基本一致的。LM-BP网络结构应用于土壤养分等级划分中,得到了很好的预测效果,为智能算法应用于农业领域奠定了良好的基础。     

基于BP神经网络的土壤容重预测模型

土壤容重是农业生产和研究的重要参数,但因成本高、工作量大等原因,其获取仍是一个紧迫的问题。数学建模技术使得科研人员尝试使用土壤传递函数PTFs间接获取土壤容重值。本研究以复合圆锥指数仪为工具,探讨应用BP神经网络建立PTFs预测土壤容重。选取粘土和粉质壤土作为实验对象,在MATLAB2008a上建立、并评价预测土壤容重的BP神经网络,即PTFs模型。研究中以均方根误差RMSE和决定系数R²为性能指标来评价所建BP神经网络。结果表明,针对复合圆锥指数仪的测量结果,应用BP神经网络算法建立PTFs可以有效预测土壤容重。粘土容重预测的决定系数R²达到0.6973,粉质壤土容重预测的决定系数R²达到0.6868。实验结果还证实土壤容重预测与测量深度无关,但与土壤类型显著相关。     

Field Test Analysis on Belled Pier Foundations under Tensile Load in Gravel Gobi

Forty full scale field tests were conducted on belled pier foundations at seven gravel Gobi sites in Gansu province and Xinjiang Uygur Autonomous Region in China. The uplift load displacement response of belled pier foundations in gravel Gobi almost followed the same pattern and presented three phase behaviors. Under tensile load, the onset of the curve usually started from the elastic linear part resulting from the compression and compaction of gravel Gobi above the enlarged base. And consequently, with the occurrence and further development of a plastic zone around the foundation, the uplift load displacement curve turned into an elasto plastic stage. Finally, the formation of whole rupture surface brought to the entire shearing damage of Gobi soil. The ultimate uplift bearing capacities were obtained by using the slope tangent method for all test foundations. Based on the theory of limiting equilibrium, the Mohr Coulomb yielding criterion and slip line field method, the circle arc rupture surface boundary condition was introduced. And the theoretical calculation equation to determine the ultimate uplift bearing capacity of the belled pier foundation in gravel Gobi was obtained. The theoretical results were compared with those of the tests and they turned out in good agreement.     

基础尺寸对碎石土地基扩底基础上拔承载力影响的现场试验研究

影响基础上拔承载能力的因素包括地基土物理力学参数及基础尺寸参数,而确定混凝土方量最小、基础上拔承载力最大的基础参数配比是基础优化设计的关键。以戈壁滩碎石土地基中的原状土扩底基础为研究对象,采用正交设计方法,以立柱直径、深宽比、扩展角为影响因素,以基础上拔承载力为分析指标,设计出9组尺寸的足尺基础。通过现场试验,获得了各试验基础的荷载位移曲线和上拔承载力值,提出了采用渐变率的概念表征荷载位移曲线的非线性变化特征,通过分析发现基础荷载位移曲线渐变率与承载能力呈负相关。结合正交试验分析结果,得出立柱直径、深宽比、扩展角3个因素中对碎石土地基原状土扩底基础抗拔承载能力的影响程度由大到小依次为深宽比、立柱直径、扩展角,表明在戈壁滩碎石土地基基础的工程设计中增加深宽比能提高基础抗拔承载能力。     

基于人工神经网络理论的土壤水分预测研究

土壤水分含量是影响作物生长的重要因素,精确的预测技术对水资源的合理利用与管理具有重要的指导意义。利用人工神经网络理论,建立了以降水量、蒸发量、相对湿度和地下水埋深为输入因子,土壤水分含量为输出因子的预测模型,并对其预测精度进行了评价。结果表明,BP神经网络模型预测土壤含水率的最大误差为8.66%,平均误差为4.27%,预测精度达到0.989。模型具有较高的预测精度,其结果可为制定合理的水资源调配方案和调度计划提供科学依据。     

Analysis of bearing capability of CFG pile composite foundation

The CFG pile is used to consolidate the foundation in one passenger dedicated line. The field experiments including low strain dynamic testing, bearing capability of single pile and the composite foundation are done, and the bearing capability of composite foundation is calculated, too. The results show that the first and second class piles account for 93.9% and 6.1% respectively, and the bearing capability eigenvalues of single pile and composite foundation are larger than the designed values. The 3D FEM model is established to simulate the stresses of CFG pile top and inter-pile soil changing with load level, and the stress ratio between CFG pile and inter-pile soil is analyzed, too. The following results are got from the calculation results of FEM: 1) the stresses of CFG pile top and inter-pile soil increase with load level, but the increasing rate of the former is larger than the latter; 2) the stress ratio between CFG pile and inter-pile soil increases quickly when the load level is low, and it tends to convergent gradually with load increasing.     

Nonlinear Forecast System for River Water Pollution Based on GIS

Based on the analysis of the water pollution spatial distribution characters of Yangtze River in Chongqing,a new method based on the integration of BP neural network and genetic arithmetic(GA) is proposed.For some shortcomings existed in the standard BP neural network,this method has ultimately overcome these shortcomings by combining the GA with BP artificial neural network through altering stimulating function,adding momentum factor to power value for BP algorithm and introducing genetic arithmetic to searching for the knots of the hidden layer,momentum factor and learning level.Using this method can easily overcome the difficulty of measuring the water prediction model's parameters.GIS is used as a tool for data management and spatial analysis,and the prediction result of the model for the water pollution spatial distribution characters of Yangtze River in Chongqing is visualized and explored with the precision of more than 78%.     

Bearing capacity of solid waste ground reinforced by stone column

Based on presented tests of improvement of solid waste ground by stone column, the failure mode of lateral expansion of column body can be found. It is assumed that solid waste surrounding column is in Rankin's positive limit situation caused by lateral expansion of column body, self weight solid waste and column is neglected, the calculation theory of limit bearing capacity of stone column can be extended on equilibrium of solid waste and column. The limit bear capacity of stone column can be calculated on parameters by complete tests and reference data. The heavy dynamic penetration tests were finished in stone column and the characteristic value of bearing capacity can be recommended on blow count. It is shown that bearing capacities of calculation and field test are close.     

基于遗传算法的BP神经网络在小麦蛋白质含量预测中的应用

为了快速、简便、准确地测定小麦蛋白质的含量,本文提出了应用近红外光谱分析技术结合遗传算法(GA)的BP神经网络的建模方法。采用SPXY算法对光谱数据进行了合理划分,并运用连续投影算法(SPA)将预处理过的数据压缩,对光谱数据提取最佳敏感波点作为GA-BP神经网络的输入,建立小麦蛋白质含量的校正模型。模型的预测均方根误差和预测相关系数为1.3379和0.979,并与BP神经网络所建立的校正模型进行了比较。结果表明：GA-BP神经网络所建模型收敛速度快、训练时间短、准确度也较高,能够实现对小麦蛋白质含量快速高效的检测。     

Fast algorithm for apparent resistivity calculation based on back propagation neural networks

To avoid the complex numerical calculation for the electromagnetic field and determine underground abnormality, a neural network based method is proposed. In consideration of turn off transmitter current, the effect of a linear ramp turn off current on transmitter is corrected. The characteristics of transient expression and the traditional calculation algorithm for apparent resistivity are analyzed, and a predigest structure of network is obtained based on the kernel expression. The three layer back propagation(BP) neural network is trained by using sample data in homogeneous half space, and its number in hidden layer was determined. The method proposed is compared with two traditional calculation methods with simulation experiments. The result demonstrates that BP neural network has a high speed of processing data and is useful in explanation of the transient electromagnetic method.     

Calculation of the Ultimate Bearing Capacity of Foundation under Triaxial Compressive Loading

It is of important significance to study the ultimate bearing capacity of foundation in three-dimensional stress. In this paper, a formula is established for ultimate bearing capacity of foundation taking three-dimensional stress into account. According to the numerical results, the ultimate bearing capacity of foundation with consideration of three-dimensional stress is improved, which is of important significance for engineering.     

Analysis on Foundation Pit Deformation of a Commercial Building in Hangzhou

The foundation pit deformation of a commercial building in Hangzhou is calculated by using calculation method related to foundation pit deformation. The calculation results by m method show that the maximum deformation is 29.2mm and the real maximum deformation is 174.6mm. The deformation character obtained by foundation coefficient method is different from the real deformation. There are lots of factors influencing foundation deformation. According to practical engineering condition, in this study, the influencing factors for soil parameters, bracing condition, condition for insertion of retaining wall into soil and construction speed are determined. The calculation method used is analyzed and modified. Finally, a useful and valid method for calculating foundation deformation is presented, by which the foundation pit deformation in practical project can be forecasted and analyzed.     

基于布谷鸟搜索神经网络的微波加热温度预测模型

微波加热是一种与被加热物直接相互作用的选择性加热方式,具有清洁、节能、减排等特点。针对工业物料作为微波加热负载时,其温度非线性变化的特点,以微波工业加热过程中的多维、海量参数为研究对象,基于泛函接神经网络模型提取样本数据的深度特征,提出了一种基于布谷鸟搜索算法,优化BP神经网络的网络参数,建立了以"数据驱动"为手段微波加热工业物料温度模型。仿真实验结果证明了所提出模型的准确性、实时性。     

神经网络自变量输入模式的视电阻率求解算法

根据瞬变电磁场理论公式中的响应和自变量之间的关系特点,提出用非线性方程模式的BP神经网络求解电阻率。通过构造单输入单输出网络结构,建立以不同时间点上的电流归一化的感应电压值为输入、视电阻率值为输出的神经网络,来拟合瞬变电磁场的二次涡流曲线。利用数值方法计算出的数据验证该训练网络的精确性,比较了不同算法对训练精度和收敛速度产生的影响。以重庆大学某处的防空洞探测实验为例验证了该算法的有效性,该算法避开具体的复杂电磁场计算或数值反问题计算,从而实现电阻率快速计算,为快速成像准备必要条件。     

A new pattern recognition method of gas sensor array detection

BP neural network based gas sensor array detection pattern recognition has some disadvantages, such as slow convergence and local minimum problem. A modified immune neural network model which combines BP algorithm and immune algorithm is proposed to enhance global search capability and improve the performance of the neural network model. Orthogonal test is adopted to design the study samples of neural network. This ensures the accuracy of neural network while reducing the number of samples. The simulation results show that the proposed pattern recognition method solves the cross sensitivity of gas sensor effectively, overcomes the disadvantages of traditional BP neural network and improves the learning speed and detection accuracy.     

基于数值预报产品的地表温度预报方法研究

为了开展地表温度预报业务,提高逐日地表温度预报准确率,利用2007—2012年的ECMWF和T213数值预报产品资料及抚顺市的逐日地表温度资料,采用逐步回归分析方法和BP神经网络模型分别构建抚顺市地表温度预报模型,并对模型的精度进行检验。结果表明,地表温度与ECMWF的高度场、海平面气压场、温度场和T213的散度场、高度场、海平面气压场、地面气压场、海平面K指数、水汽通量、相对湿度、温度场、地面气温和场涡度场均呈显著相关。对预报模型进行精度检验显示,地表平均温度和地表最低温度的预报效果较好,≤3℃预报准确率均达到79%以上。2种模型对比显示,BP神经网络预报模型总体上优于逐步回归预报模型;逐步回归预报模型较BP神经网络预报模型稳定。     

联想神经网络的风速序列预测分析

为了提高风速序列预测的可靠性,针对具有混沌特性的风速序列,构造了一种用于风速序列预测的联想网络。以风速序列的波动性作为相似性测度准则,构造联想网络的存储样本模式,根据存储模式中蕴含的关联信息完成网络的无监督学习,从而完成具有自相似性的风速序列的一步或多步预测分析。与传统前向型神经网络相比,该网络预测机理明确,预测结果唯一,且可一次给出多步预测结果。仿真实验结果表明,该网络的具有良好预测性能,适用于风速序列的动态预测。     

BP神经网络在许昌土壤墒情预测模型的应用

为了科学指导农业灌溉和抗旱救灾,通过采用BP神经网络方法,研究许昌现有土壤墒情经验分析模型,建立了许昌土壤墒情预测模型并结合许昌多年的土壤水分实测数据和气象资料进行误差分析,应用结果表明BP神经网络模型可较好地应用于许昌土壤墒情分析和预测,对于不同条件的地区具有较好的适应性和推广应用价值。     

罗非鱼池塘养殖溶解氧预测研究

为及时有效地掌握池塘养殖中溶解氧浓度的变化趋势,保障罗非鱼稳定高效养殖,在对罗非鱼池塘养殖实际情况进行研究和分析的基础上,采用粒子群算法对BP神经网络模型进行参数优化,针对无锡市2015 年8 月23—11 月4 日这段时间内南泉实验基地的水产养殖溶解氧进行预测。同时,将粒子群优化BP神经网络模型与BP神经网络模型的训练结果和预测结果进行对比。研究结果表明,PSO-BP优化模型的训练和预测结果远远优于普通BP神经网络模型,除异常点外,误差率基本均低于0.5%。同时,该模型收敛速度快,计算复杂性低,能够较好的体现和预测罗非鱼池塘养殖的溶解氧趋势,也为其他水质指标的预测提供了研究方向。     

artificial neural networks based prediction of electromagnetic compatibility problems

It is necessary to predict electromagnetic compatibility (EMC) for electronic equipment and systems. We proposed a fast EMC prediction approach via artificial neural networks (ANN). By choosing relevant electromagnetic interference parameters as the input prediction features, a back propagation (BP) neural network was used to construct the mapping between the input prediction features and the electromagnetic disturbance response of the sensitive system. The EMC fast prediction BP model was trained and tested by sample sets generated using an electromagnetic computational method. We used this method to predict the crosstalk coupling between two wires. The experimental results show the effectiveness of the proposed method.