A statistical reappraisal of the relationship between liquid limit and specific surface area,cation exchange capacity and activity of clays

More than 500 datasets from the literature have been used to evaluate the relationships of specific surface area(SSA),cation exchange capacity(CEC)and activity versus the liquid limit(LL).The correlations gave R~2 values ranging between 0.71 and 0.92.Independent data were also used to validate the correlations.Estimated SSA values slightly overestimate the measured SSA up to 100 m~2/g.Regarding the estimated CEC values,they overestimated the measured CEC values up to 20 meq/(100 g).A probabilistic approach was performed for the correlations of SSA,CEC and activity versus LL.The analysis shows that the relations of SSA,CEC and activity with LL are robust.Using the LL values,it is possible to assess other basic engineering properties of clays.     

Prediction of maximum surface settlement caused by earth pressure balance (EPB) shield tunneling with ANN methods

In order to determine the appropriate model for predicting the maximum surface settlement caused by EPB shield tunneling, three artificial neural network (ANN) methods, back-propagation (BP) neural network, the radial basis function (RBF) neural network, and the general regression neural network (GRNN), were employed and the results were compared. The nonlinear relationship between maximum ground surface settlements and geometry, geological conditions, and shield operation parameters were considered in the ANN models. A total number of 200 data sets obtained from the Changsha metro line 4 project were used to train and validate the ANN models. A modified index that defines the physical significance of the input parameters was proposed to quantify the geological parameters, which improves the prediction accuracy of ANN models. Based on the analysis, the GRNN model was found to outperform the BP and RBF neural networks in terms of accuracy and computational time. Analysis results also indicated that strong correlations were established between the predicted and measured settlements in GRNN model with MAE = 1.10, and RMSE = 1.35, respectively. Error analysis revealed that it is necessary to update datasets during EPB shield tunneling, though the database is huge.     

Prediction of railway track geometry deterioration using artificial neural networks: a case study for Turkish state railways

The main goal of this paper is to model track geometry deterioration using a comprehensive field investigation gathered over a period of 2 years on approximately 180 km of railway line. Artificial neural networks (ANNs) were adapted for this research. The railway line was divided into analytical segments (ASs). For each AS, the following data were collected: track structure, traffic characteristics, track layout, environmental factors, track geometry, and maintenance and renewal data. ANN models were developed for the main track geometry parameters and produced significant relationships between the variables. In addition, sensitivity analyses were performed to compute the importance of each predictor in determining the neural network. The obtained results proved that ANN may be an alternative method for predicting track geometry deterioration.     

Prediction of rubberized concrete properties using artificial neural network and fuzzy logic

In this study, waste automobile tyres in two different sizes were used in production of rubberized fresh concretes. Their unit weight and flow table values were determined experimentally. The values determined were also found when artificial neural networks (ANN) and fuzzy logic (FL) models were employed. According to the given rubberized concrete data, it was demonstrated that properties of fresh concrete could be determined without attempting any experiments by using ANN and FL models. During the tests similar results were observed for experimental results with those of ANN and FL models. Besides, the facts that lighter concrete might be produced using tyre as a light material and waste tyres may be recycled this way were put forth.     

高吸力段黏土水合机制及微观持水模型研究

采用水汽吸附法测定3种蒙脱土在高吸力段的持水特征曲线,基于X射线衍射及BET吸附理论架构,提出了两类水合概化模型并建立了阳离子交换量、比表面积等参数的计算方法。基于极低相对湿度范围段(RH0.15)持水能力只受控于层间阳离子水合的机理认识,通过阳离子与水分子相互作用能量方程,推导了极高吸力下(ψ250 MPa)持水曲线微观参数模型。研究表明:层间阳离子水合能力的差异会导致蒙脱土吸附起始阶段呈现不同趋势,对于低水合能阳离子交换土,水分子首先吸附于黏土颗粒外表面,之后随相对湿度增加逐渐进入层间吸附,反之,则直接进入层间离子水合阶段;基于BET曲线计算的阳离子交换量、比表面积值与实测值之间吻合较好,构建的持水模型能够对文献报道数据进行有效预测,该模型可量化表征阳离子交换量、化合价、离子半价等微观参数对吸力势的影响程度。     

Simulation-based construction productivity forecast using Neural-Network-Driven Fuzzy Reasoning

Fuzzy models and Artificial Neural Network (ANN) systems are two well-known areas of soft-computing that have significantly helped researchers with decision-making under uncertainties. Uncertainty, an ever-present factor in construction projects, has made such intelligent systems very attractive to the construction industry. Estimating the productivity of construction operations, as a basic element of project planning and control, has become a remarkable target for forecasting models. A glimpse into this interdisciplinary field of research exposes the need for a system, that (1) models the effect of qualitative and quantitative variables on construction productivity with an improved accuracy of estimation and (2) has the ability to deal with both crisp and fuzzy input variables in one single framework. Neural-Network-Driven Fuzzy Reasoning (NNDFR), as one of the hybrid intelligent structures, displays a great potential for modeling datasets among which clear clusters are recognizable. The weakness of NNDFR in auto-tuning the design of fuzzy membership functions along with this model's insufficient attention to the optimization of number of clusters has created an area for further research. In this paper, the parameters (fuzzifier and number of clusters) of the proposed system are optimized by using Genetic Algorithm (GA) to fine-tune the system for the highest possible level of accuracy that can be exploited for productivity estimation. The proposed model is also capable of dealing with a combination of crisp and fuzzy input variables by using a hybrid modeling approach based on the application of the alpha-cut technique. The developed model helps researchers and practitioners use historical data to forecast the productivity of construction operations with a level of accuracy greater than what could be offered by traditional techniques.     

Training Artificial Neural Networks with the Aid of Fuzzy Sets

This article presents a technique of training artificial neural networks (ANNs) with the aid of fuzzy sets theory. The proposed ANN model is trained with field observation data for predicting the collapse potential of soils. This ANN model uses seven soil parameters as input variables. The output variable is the collapsibility (whether the soil is collapsible) or the collapse potential (if the soil is judged collapsible). The proposed technique involves a module for preprocessing input soil parameters and a module for postprocessing network output. The preprocessing module screens the input data through a group of predefined fuzzy sets, and the postprocessing module, on the other hand, "defuzzifies" the output from the network into a "nonfuzzy" collapse potential, a single value. The ANN with the proposed preprocessing and post-process techniques is shown to be superior to the conventional ANN model in the present study.     

Application of artificial neural network for the prediction of jaggery mass during drying inside the natural convection greenhouse dryer

In this paper, an attempt is made to predict the hourly mass of jaggery during the process of drying inside greenhouse dryer under the natural convection mode. Jaggery was dried until the constant variation in the mass of jaggery. Artificial neural network (ANN) is used to predict the mass of the dried jaggery on hourly basis. Solar radiation, ambient temperature and relative humidity are input parameters for the prediction of jaggery mass in each hour in the ANN modelling. The results of the ANN model are also validated with experimental drying data of jaggery mass. The statistical parameters such as root mean square error and correlation coefficient (R²) are used to measure the difference between values predicted by the ANN model and the values actually observed from the experimental study. It was found that the results of the ANN model and experimental are shown fairly good agreement.     

民用建筑供热负荷的神经网络法预测

分析了供热系统负荷变化的各种扰量,提出利用人工神经网络对供热负荷进行预测的方法。对神经网络预测的可行性、方法的实施内容及输入输同变量的选择,网络连接方法的选择等进行了讨论。在进一步对供热负荷特性研究的基础上,可以利用人工神经网络对其进行切实可行的预测。     

Exploring the possibility of assessing the damage degree of liquefaction based only on seismic records by artificial neural networks

This study presents a new approach to determine the damage degree of liquefaction caused by a large earthquake. We propose an artificial neural network (ANN) model based only on the seismic records of ground and define the degree of liquefaction “DDL” as a damage index. This ANN model predicts the degree of excess pore water pressure increase as the correct output label based on the seismic records obtained from the three-dimensional shaking table test. The proposed model achieved high accuracy, and the outcomes from training data indicated that the ANN model is suitable to function as a liquefaction assessment system. Further, to evaluate the applicability of the proposed ANN model in the real world, the datasets of waves from three actual seismic records were input to the ANN as validation data. The DDL judgment obtained was a good fit with the real phenomena observed.     

A cost-effective chemo-thermo-poroelastic wellbore stability model for mud weight design during drilling through shale formations

Drilling through chemically-active shale formations is of special importance due to time-dependent drilling fluid-shale interactions.The physical models presented so far include sophisticated input parameters,requiring advanced experimental facilities,which are costly and in most cases unavailable.In this paper,sufficiently-accurate,yet highly practical,models are presented containing parameters easilyderived from well-known data sources.For ion diffusivity coefficient,the chemical potential was formulated based on the functionality of water activity to solute concentration for common solute species in field.The reflection coefficient and solute diffusion coefficient within shale membrane were predicted and compared with experimental measurements.For thermally-induced fluid flow,a model was utilized to predict thermo-osmosis coefficient based on the energy of hydrogen-bond that attained a reasonably-accurate estimation from petrophysical data,e.g.porosity,specific surface area(SSA),and cation exchange capacity(CEC),The coupled chemo-thermo-poroelastic governing equations were developed and solved using an implicit finite difference scheme.Mogi-Coulomb failure criterion was adopted for mud weight required to avoid compressive shear failure and a tensile cut-off failure index for mud weight required to prevent tensile fracturing.Results showed a close agreement between the suggested model and experimental data from pressure transmission tests.Results from a numerical example for a vertical wellbore indicated that failure in shale formations was time-dependent and a failure at wellbore wall after 85 min of mud-shale interactions was predicted.It was concluded that instability might not firstly occur at wellbore wall as most of the conventional elastic models predict;perhaps it occurs at other points inside the formation.The effect of the temperature gradient between wellbore and formation on limits of mud window confirmed that the upper limit was more sensitive to the temperature gradient than the lower limit.     

Application of artificial neural networks for predicting the cuttability of rocks by drag tools

Many models have previously been developed for predicting specific cutting energy (SE), being the measure of rock cuttability, from intact rock properties employing conventional multiple linear or nonlinear regression techniques. Artificial neural networks (ANN) also have a great potential in building such models. This paper is concerned with the application of ANN for the prediction of cuttability of rocks from their intact properties. For that purpose, data obtained from three different projects were subjected to statistical analyses using MATLAB. Principal components analysis together with the scatterplots of SE against intact rock properties were employed to select the predictors for SE models. Results of the principal components analysis have shown that the most of the variance in the data set can be explained by three principal components. Principal component with the highest variance is weighted mainly on the uniaxial compressive strength (UCS), Brazilian tensile strength (BTS), static modulus of elasticity (Elasticity), and cone indenter hardness (CI), which were regarded as the independent variables driving the data set. Three predictive models for SE were developed employing above independent variables by multiple nonlinear regression with forward stepwise method and ANN, respectively. Neural networks were developed for two different numbers of hidden neurons in the hidden layer. Goodness of the fit measures revealed that ANN models fitted the data as accurately as multiple nonlinear regression model, indicating the usefulness of artificial neural networks in predicting rock cuttability.     

Fuzzy Cellular Automata Model for Signalized Intersections

At signalized intersections, the decision‐making process of each individual driver is a very complex process that involves many factors. In this article, a fuzzy cellular automata (FCA) model, which incorporates traditional cellular automata (CA) and fuzzy logic (FL), is developed to simulate the decision‐making process and estimate the effect of driving behavior on traffic performance. Different from existing models and applications, the proposed FCA model utilizes fuzzy interface systems (FISs) and membership functions to simulate the cognition system of individual drivers. Four FISs are defined for each decision‐making process: car‐following, lane‐changing, amber‐running, and right‐turn filtering. A field observation study is conducted to calibrate membership functions of input factors, model parameters, and to validate the proposed FCA model. Simulation experiments of a two‐lane system show that the proposed FCA model is able to replicate decision‐making processes and estimate the effect on overall traffic performance.     

Predicting the strength properties of slurry infiltrated fibrous concrete using artificial neural network

This paper is aimed at adapting Artificial Neural Networks (ANN) to predict the strength properties of SIFCON containing different minerals admixture. The investigations were done on 84 SIFCON mixes, and specimens were cast and tested after 28 days curing. The obtained experimental data are trained using ANN which consists of 4 input parameters like Percentage of fiber (PF), Aspect Ratio (AR), Type of admixture (TA) and Percentage of admixture (PA). The corresponding output parameters are compressive strength, tensile strength and flexural strength. The predicted values obtained using ANN show a good correlation between the experimental data. The performance of the 4-14-3 architecture was better than other architectures. It is concluded that ANN is a highly powerful tool suitable for assessing the strength characteristics of SIFCON.     

Hybrid intelligent inference model for enhancing prediction accuracy of scour depth around bridge piers

Bridge-pier scouring is a main cause of bridge failures. Thus, accurately predicting the scour depth around bridge piers is critical, both to specify adequate depths for new bridge foundations and to assess/monitor the safety of existing bridges. This study proposes a novel artificial intelligence (AI) model, the intelligent fuzzy radial basis function neural network inference model (IFRIM), to estimate future scour depth around bridge piers. IFRIM is a hybrid of the radial basis function neural network (RBFNN), fuzzy logic (FL), and the artificial bee Cclony (ABC) algorithm. In the IFRIM, FL is used to handle the uncertainties in input information, RBFNN is used to handle the fuzzy input–output mapping relationships, and the ABC search engine employs optimisation to identify the most suitable tuning parameters for RBFNN and FL based on minimal error estimation. A 10-fold cross-validation method finds that the IFRIM model achieves at least 21% and 14.5% reductions in root mean square error and mean absolute error values, respectively, compared with other AI techniques. Study results support the IFRIM as a promising new tool for civil engineers to predict future scour depth around bridge piers.     

Prediction of blast-induced ground vibration using artificial neural networks

Blasting is still being considered to be one the most important applicable alternatives for conventional tunneling. Ground vibration generated due to blasting is an undesirable phenomenon which is harmful for the nearby habitants and dwellings and should be prevented. In this paper, an attempt has been made to predict blast-induced ground vibration using artificial neural network (ANN) in the Siahbisheh project, Iran. To construct the model maximum charge per delay, distance from blasting face to the monitoring point, stemming and hole depth are taken as input parameters, whereas, peak particle velocity (PPV) is considered as an output parameter. A database consisting of 182 datasets was collected at different strategic and vulnerable locations in and around the project. From the prepared database, 162 datasets were used for the training and testing of the network, whereas 20 randomly selected datasets were used for the validation of the ANN model. A four layer feed-forward back-propagation neural network with topology 4-10-5-1 was found to be optimum. To compare performance of the ANN model with empirical predictors as well as regression analysis, the same database was applied. Superiority of the proposed ANN model over empirical predictors and statistical model was examined by calculating coefficient of determination for predicted and measured PPV. Sensitivity analysis was also performed to get the influence of each parameter on PPV. It was found that distance from blasting face is the most effective and stemming is the least effective parameter on the PPV.     

An artificial neural network-based mathematical model for the prediction of blast-induced ground vibrations

ABSTRACT

This paper presents an artificial neural network (ANN) based mathematical model for the prediction of blast-induced ground vibrations using the data obtained from the literature. A feed-forward back-propagation multi-layer perceptron (MLP) was adopted, and the Levenberg–Marquardt algorithm was used in training the network. The powder factor, the maximum charge per delay, and distance from blasting face to monitoring point are the input variables. The peak particle velocity (PPV) is the targeted output variable. The model was then formulated using the weights and biases output from the ANN simulation. Multilinear regression (MLR) analysis was also performed using the same number of datasets, as in the case of ANN. The quality of the proposed ANN-based model was also tested with another 14 datasets outside the one used in developing the models and compared with more classical models. The coefficient of the determination (R²) of the proposed ANN-based model was the highest.     

Backfilling missing microbial concentrations in a riverine database using artificial neural networks

Predicting peak pathogen loadings can provide a basis for watershed and water treatment plant management decisions that can minimize microbial risk to the public from contact or ingestion. Artificial neural network models (ANN) have been successfully applied to the complex problem of predicting peak pathogen loadings in surface waters. However, these data-driven models require substantial, multiparameter databases upon which to train, and missing input values for pathogen indicators must often be estimated. In this study, ANN models were evaluated for backfilling values for individual observations of indicator bacterial concentrations in a river from 44 other related physical, chemical, and bacteriological data contained in a multi-year database. The ANN modeling approach provided slightly superior predictions of actual microbial concentrations when compared to conventional imputation and multiple linear regression models. The ANN model provided excellent classification of 300 randomly selected, individual data observations into two defined ranges for fecal coliform concentrations with 97% overall accuracy. The application of the relative strength effect (RSE) concept for selection of input variables for ANN modeling and an approach for identifying anomalous data observations utilizing cross validation with ANN model are also presented.     

An advanced ANN model for predicting the rotational behaviour of semi-rigid composite joints in fire using the back-propagation paradigm

This paper describes an artificial neural networking (ANN) model developed to predict the behaviour of semi-rigid composite joints at elevated temperature. Three different semi-rigid composite joints were selected, two flexible end-plates and one flush end-plate. Seventeen different parameters were selected as input parameters representing the geometrical and mechanical properties of the joints as well as the joint’s temperature and the applied loading, and used to model the rotational capacity of the joints with increasing temperatures. Data from experimental fire tests were used for training and testing the ANN model. Results from nine experimental fire tests were evaluated with a total of 280 experimental cases. The results showed that the R² value for the training and testing sets were 0.998 and 0.97, respectively. This indicates that results from the ANN model compared well with the experimental results demonstrating the capability of the ANN simulation techniques in predicting the behaviour of semi-rigid composite joints in fire. The described model can be modified to study other important parameters that can have considerable effect on the behaviour of joints at elevated temperatures such as temperature gradient, axial restraints, etc.     

The unique relationship between swell percent and swell pressure of compacted clays

Expansive soils exhibit high volumetric deformations, posing a serious threat to the stability of structures and foundations. However, measurement of swelling properties is time consuming and requires special and expensive equipment. This study made an attempt to investigate the relationship between these parameters and easily obtained soil properties using various clay mineral mixtures to obtain soils in a wide range of plasticity indices. Free swell percent was correlated to clay percent, water content, dry unit weight, plasticity index, liquidity index and cation exchange capacity using multiple regression analyses. A very high (R = 0.94) fit was also found for a proposed relationship between the percent swell and swell pressure values for samples having a swell pressure ≤300 kPa. It is concluded that the proposed equations offer a rapid and inexpensive substitute for laboratory testing of swell percent/swell pressure in the preliminary stages of site investigations.