Interpretation of concrete dam behaviour with artificial neural network and multiple linear regression models

The safety control of large dams is based on the measurement of some important quantities that characterize their behaviour (like absolute and relative displacements, strains and stresses in the concrete, discharges through the foundation, etc.) and on visual inspections of the structures. In the more important dams, the analysis of the measured data and their comparison with results of mathematical or physical models is determinant in the structural safety assessment.In its lifetime, a dam can be exposed to significant water level variations and seasonal environmental temperature changes. The use of statistical models, such as multiple linear regression (MLR) models, in the analysis of a structural dam’s behaviour has been well known in dam engineering since the 1950s. Nowadays, artificial neural network (NN) models can also contribute in characterizing the normal structural behaviour for the actions to which the structure is subject using the past history of the structural behaviour. In this work, one important aspect of NN models is discussed: the parallel processing of the information.This study shows a comparison between MLR and NN models for the characterization of dam behaviour under environment loads. As an example, the horizontal displacement recorded by a pendulum is studied in a large Portuguese arch dam. The results of this study show that NN models can be a powerful tool to be included in assessments of existing concrete dam behaviour.     

Physical and neural network models of a silica-gel desiccant wheel

Two methods for modelling the performance of a desiccant wheel are presented: a physical model, based on mass and energy balances of the process, and a neural network model, based on the training of a black box model with real data. The physical model consists of a set of non-linear differential equations solved by finite differences techniques. The neural network model consists of a four-input–four-output network that calculates the outlet conditions from inlet ones. Real data are used to validate the physical model and to train the neural network. The physical model shows discrepancies between calculated and measured values mainly due to the fact that the system is assumed to be adiabatic. The heat losses in the ducts and the wheel are not considered in the model, but in the experimental facility these losses occur. In the case of the neural network model, the temperature and humidity ratio calculated for the outlet air are in accordance with the experimental data.     

Estimation and analysis of preference heterogeneity in residential choice behaviour

Effective and efficient planning and development of residential environments require clarifying the nature of residential preferences. In reality, residential preferences are heterogeneous, so the standard econometric models that assume only one type of preference are not optimal. In this study, conjoint choice experiment methods are employed with a mixed logit approach. The findings reveal significant heterogeneity with regard to some residential attributes. The determinants of preference heterogeneity were also investigated by conducting regression analyses on the attributes that were valued heterogeneously. Overall, the relationships observed between the explanatory variables and the heterogeneity in the valuations were understandable. However, coefficient of determination values for each model were low, indicating that the bulk of preference heterogeneity results from unobservable factors.     

Airfield pavement deterioration assessment using stress-dependent neural network models

In this study, an artificial neural network (ANN)-based approach was employed to backcalculate the asphalt concrete and non-linear stress-dependent subgrade moduli from non-destructive test (NDT) data acquired at the Federal Aviation Administration's National Airport Pavement Test Facility (NAPTF) during full-scale traffic testing. The ANN models were trained with results from an axisymmetric finite element pavement structural model. Using the ANN-predicted moduli based on the NDT test results, the relative severity effects of simulated Boeing 777 (B777) and Boeing 747 (B747) aircraft gear trafficking on the structural deterioration of NAPTF flexible pavement test sections were characterized. The results indicate the potential of using lower force amplitude NDT test data for routine airport pavement structural evaluation, as long as they generate sufficient deflections for reliable data acquisition. Therefore, NDT tests that employ force amplitudes at prototypical aircraft loading may not be necessary to evaluate airport pavements.     

Intelligent modelling of sandstone deformation behaviour using fuzzy logic and neural network systems

Bulletin of Engineering Geology and the Environment - A realistic analysis of rock deformation in response to any change in stresses is heavily dependent on the reliable determination of the rock...     

Prediction of bed load sediments using different artificial neural network models

Modeling and prediction of bed loads is an important but difficult issue in river engineering. The introduced empirical equations due to restricted applicability even in similar conditions provide different accuracies with each other and measured data. In this paper, three different artificial neural networks (ANNs) including multilayer percepterons, radial based function (RBF), and generalized feed forward neural network using five dominant parameters of bed load transport formulas for the Main Fork Red River in Idaho-USA were developed. The optimum models were found through 102 data sets of flow discharge, flow velocity, water surface slopes, flow depth, and mean grain size. The deficiency of empirical equations for this river by conducted comparison between measured and predicted values was approved where the ANN models presented more consistence and closer estimation to observed data. The coefficient of determination between measured and predicted values for empirical equations varied from 0.10 to 0.21 against the 0.93 to 0.98 in ANN models. The accuracy performance of all models was evaluated and interpreted using different statistical error criteria, analytical graphs and confusion matrixes. Although the ANN models predicted compatible outputs but the RBF with 79% correct classification rate corresponding to 0.191 network error was outperform than others.     

基于BP神经网络的管网维修管径预测分析

以MH水司2004-2012年供水管网维修数据作为研究对象,以BP神经网络模型为研究方法,构建了MH水司供水管网维修的预测模型,对供水管网中待维修管道和管件的管径分布作了短期趋势预测。预测结果表明,该模型的预测精度较高,平均偏差最大为0．0054,均方差最大为0．0077;并给出了DN≤50、50〈DN≤100、100〈DN≤150、150〈DN≤200、200〈DN≤300、300〈DN≤500、500〈DN≤800和800〈DN≤1600的管道维修数量在历年和年内管道维修记录统计分析结果中的变化规律。     

Multiscale computation on feedforward neural network and recurrent neural network

Homogenization methods can be used to predict the effective macroscopic properties of materials that are heterogenous at micro- or fine-scale. Among existing methods for homogenization, computational homogenization is widely used in multiscale analyses of structures and materials. Conventional computational homogenization suffers from long computing times, which substantially limits its application in analyzing engineering problems. The neural networks can be used to construct fully decoupled approaches in nonlinear multiscale methods by mapping macroscopic loading and microscopic response. Computational homogenization methods for nonlinear material and implementation of offline multiscale computation are studied to generate data set. This article intends to model the multiscale constitution using feedforward neural network (FNN) and recurrent neural network (RNN), and appropriate set of loading paths are selected to effectively predict the materials behavior along unknown paths. Applications to two-dimensional multiscale analysis are tested and discussed in detail.     

Consumer choice and optimal locations models: Formulations and heuristics

A new direction of research in competitive location theory incorporates theories of consumer choice behavior in its models. Following this direction, the present article studies the importance of consumer behavior with respect to distance or transportation costs in the optimality of locations obtained by traditional competitive location models. We consider various ways of defining a key parameter in the basic maximum capture model (MAXCAP). This parameter will indicate a number of ways to take distance into account based on several consumer choice behavior theories. The optimal locations and the deviation in demand – captured when the optimal locations of the other models are used instead of the true ones – are computed for each model. We present a metaheuristic based on GRASP and the tabu search procedure to solve all the models. Computational experience and an application to a 55-node network are also presented. Received: 12 June 1998 / Accepted: 29 July 1999     

Black-white regional migration and the impact of education: a multinomial logit analysis

This study investigates the demographic determinants of regional migration within the United States for the periods 1965–70 and 1975–80. Primary focus of the paper concerns racial differences in an individual's choice of destination region and the influence of educational attainment on this decision. Data used in the analysis are drawn from the United States census public-use microdata samples. Strong evidence is provided demonstrating that an individual's choice of destination region differs by race and that educational attainment has a significant impact on racial differences in the direction of regional migration. Moreover, racial selectivity and the influence of education is shown to be significantly different between time periods.I am grateful to Michael J. Greenwood and an anonymous referee for insightful comments.     

Slope stability analysis based on big data and convolutional neural network

The Limit Equilibrium Method (LEM) is commonly used in traditional slope stability analyses, but it is time-consuming and complicated. Due to its complexity and nonlinearity involved in the evaluation process, it cannot provide a quick stability estimation when facing a large number of slopes. In this case, the convolutional neural network (CNN) provides a better alternative. A CNN model can process data quickly and complete a large amount of data analysis in a specific situation, while it needs a large number of training samples. It is difficult to get enough slope data samples in practical engineering. This study proposes a slope database generation method based on the LEM. Samples were amplified from 40 typical slopes, and a sample database consisting of 20000 slope samples was established. The sample database for slopes covered a wide range of slope geometries and soil layers’ physical and mechanical properties. The CNN trained with this sample database was then applied to the stability prediction of 15 real slopes to test the accuracy of the CNN model. The results show that the slope stability prediction method based on the CNN does not need complex calculation but only needs to provide the slope coordinate information and physical and mechanical parameters of the soil layers, and it can quickly obtain the safety factor and stability state of the slopes. Moreover, the prediction accuracy of the CNN trained by the sample database for slope stability analysis reaches more than 99%, and the comparisons with the BP neural network show that the CNN has significant superiority in slope stability evaluation. Therefore, the CNN can predict the safety factor of real slopes. In particular, the combination of typical actual slopes and generated slope data provides enough training and testing samples for the CNN, which improves the prediction speed and practicability of the CNN-based evaluation method in engineering practice.     

边坡岩体稳定性的人工神经网络预测模型

在综合分析边坡岩体变形失稳破坏模式及其影响因素的基础上 ,提出了表征边坡岩体稳定性分析的复合指标。以大量水电边坡工程的稳定状况为学习训练样本及预测样本 ,讨论了基于人工神经网络技术的边坡岩体稳定性分析方法及其有效性。研究表明 ,用人工神经网络方法预测边坡岩体的稳定状况是可行的。     

Disinfection by-product formation following chlorination of drinking water: Artificial neural network models and changes in speciation with treatment

Artificial neural network (ANN) models were developed to predict disinfection by-product (DBP) formation during municipal drinking water treatment using the Information Collection Rule Treatment Studies database complied by the United States Environmental Protection Agency. The formation of trihalomethanes (THMs), haloacetic acids (HAAs), and total organic halide (TOX) upon chlorination of untreated water, and after conventional treatment, granular activated carbon treatment, and nanofiltration were quantified using ANNs. Highly accurate predictions of DBP concentrations were possible using physically meaningful water quality parameters as ANN inputs including dissolved organic carbon (DOC) concentration, ultraviolet absorbance at 254 nm and one cm path length (UV₂₅₄), bromide ion concentration (Br⁻), chlorine dose, chlorination pH, contact time, and reaction temperature. This highlights the ability of ANNs to closely capture the highly complex and non-linear relationships underlying DBP formation. Accurate simulations suggest the potential use of ANNs for process control and optimization, comparison of treatment alternatives for DBP control prior to piloting, and even to reduce the number of experiments to evaluate water quality variations when operating conditions are changed. Changes in THM and HAA speciation and bromine substitution patterns following treatment are also discussed.     

基于神经网络的结构损伤识别综述

简单介绍了神经网络技术及其分类方法,对使用神经网络进行斜拉桥损伤识别的基本流程进行了详细阐述,并分析了输入向量的选择优缺点,以期促进基于神经网络的结构损伤识别技术的推广应用。     

基于参数相关性和混合神经网络的结构随机灵敏度分析方法

针对进行随机分析时采用蒙特卡罗计算法效率低,未能考虑参数之间相关性,导致在分析参变量对结构力学性能的影响时得到错误的灵敏度系数,以及原有灵敏度计算方法只能考虑局部梯度等问题,采用改进的混沌粒子群算法优化网络寻址结构,利用混合神经网络构建复杂结构响应的近似模型,通过相关参数与独立正态参数之间的等效变换建立符合参数相关性的随机序列对结构进行随机性分析,并根据文中提出的灵敏度度量方法计算随机变量的全局灵敏度系数。通过算例验证所提方法的可行性,且考虑参数之间相关关系得到的结构随机响应更符合工程实际情况。同时,利用所提出的随机灵敏度计算方法可以更好地反映各随机变量对结构响应的相关性和敏感性。     

无腹筋钢筋混凝土简支梁受剪特性的神经网络模型初探

采用人工神经网络方法，分别探讨了均布荷载及两点加载下无腹筋钢筋混凝土简支梁的斜截面抗剪特性，建立了相应的模型，寻找到两种加载形式下钢筋混凝土简支梁的各种输人参数与实际抗剪承载力之间的关系。通过原始实验结果和理论模型分析结果的对比，表明了该模型的可行性及适用性。该模型的建立，可为我国的抗剪理论提供参考。     

Prediction of the compressive strength of no-slump concrete: A comparative study of regression,neural network and ANFIS models

No-slump concrete (NSC) is defined as concrete having either very low or zero slump that traditionally used for prefabrication purposes. The sensitivity of NSC to its constituents, mixture proportion, compaction, etc., enforce some difficulties in the prediction of the compressive strength. In this paper, by considering concrete constituents as input variables, several regression, neural networks (NNT) and ANFIS models are constructed, trained and tested to predict the 28-days compressive strength of no-slump concrete (28-CSNSC). Comparing the results indicate that NNT and ANFIS models are more feasible in predicting the 28-CSNSC than the proposed traditional regression models.     

Elastic structural analysis based on graph neural network without labeled data

Artificial intelligence is gaining increasing popularity in structural analysis. However, at the structural system level, the appropriateness of data representation, the paucity of data, and the physical interpretability of results are rarely studied and remain profound challenges. To fill such gaps, a physics-informed model named StructGNN-E (i.e., structural analysis based on graph neural network [GNN]–elastic) based on the GNN architecture, which is capable of implementing the elastic analysis of structural systems without labeled data, is proposed in this study. The systems with structural topologies and member configurations are organized as graph data and later processed by a modified graph isomorphism network. Moreover, to avoid dependence on big data, a novel physics-informed paradigm is proposed to incorporate mechanics into deep learning (DL), ensuring the theoretical correctness of the results. Numerical experiments and ablation studies demonstrate the unique effectiveness of StructGNN-E against common DL models, with an average accuracy of 99% and excellent computational efficiency. Due to its differentiability, StructGNN-E is promising for bidirectionally linking structural parameters and analysis results, paving the way for a new end-to-end structural optimization method in the future.     

Structural reliability analysis for implicit performance functions using artificial neural network

The Monte-Carlo simulation (MCS), the first-order reliability methods (FORM) and the second-order reliability methods (SORM), are three reliability analysis methods that are commonly used for structural safety evaluation. The MCS requires the calculations of hundreds and thousands of performance function values. The FORM and SORM demand the values and partial derivatives of the performance function with respect to the design random variables. Such calculations could be time-consuming or cumbersome when the performance functions are implicit. Such implicit performance functions are normally encountered when the structural systems are complicated and numerical analysis such as finite element methods has to be adopted for the prediction.To address this issue, this paper presents three artificial neural network (ANN)-based reliability analysis methods, i.e. ANN-based MCS, ANN-based FORM, and ANN-based SORM. These methods employ multi-layer feedforward ANN technique to approximate the implicit performance functions. The ANN technique uses a small set of the actual values of the implicit performance functions. Such a small set of actual data is obtained via normal numerical analysis such as finite element methods for the complicated structural system. They are used to develop a trained ANN generalization algorithm. Then a large number of the values and partial derivatives of the implicit performance functions can be obtained for conventional reliability analysis using MCS, FORM or SORM. Examples are given in the paper to illustrate why and how the proposed ANN-based structural reliability analysis can be carried out. The results have shown the proposed approach is applicable to structural reliability analysis involving implicit performance functions. The present results are compared well with those obtained by the conventional reliability methods such as the direct Monte-Carlo simulation, the response surface method and the FORM method 2.