Search

Regular use of artificial neural networks (ANN) analysis for predicting the vertical swelling percentage of expansive clays may lead to inappropriate results in terms of their geophysical behavior. This paper presents two new ANN Models derived from a two-stage procedure. The models were estimated using the same data set from the previous paper, and their statistical fit was clearly found to be superior in comparison to the previous models. Furthermore, one of these two models exhibited the expected geophysical behavior. As this new ANN Model yields higher predicted swelling-percentage values, it can definitely be regarded as a preferable one in the sense of enlarging the safety margin in heave calculations. [ABSTRACT FROM AUTHOR]

Artificial neural networks (ANNs) are generally regarded to behave as black-box systems. Recent research explores various methods that can provide an insight into the internal connections and relationships existing within the network. Various methodologies that understand the input variable contribution are analyzed in detail, and rule extraction approaches for a trained artificial neural network are addressed. To understand the contribution of input variables to rainfall-runoff model combination systems, this paper for the first time investigates knowledge extraction from artificial neural network, which is used to combine the results obtained from different competing rainfall-runoff models, using three different approaches: (1) Garson's algorithm; (2) neural interpretation diagram (NID); and (3) sensitivity analysis (SA). For the purpose of investigating knowledge extraction techniques, the trained multilayer perceptron neural network to combine the results from four different rainfall-runoff models for the Brosna Catchment located in Ireland has been chosen. The results of the three approaches obtained in this study indicate that they can be used to reduce the complexity of rainfall-runoff model combination systems by eliminating the least significant contributing input variables. Based on these approaches, the paper helps to provide guidance in the optimal number of rainfall-runoff models that best perform in a combination system. [ABSTRACT FROM AUTHOR]

In this paper, a neuro-proportional derivative (PD) method is presented to stabilize a satellite with respect to the gravity gradient. Three reaction wheels are employed to produce the necessary torque in the axes of the satellite. Quaternion and angular velocity vectors are used in the PD controller, and a neural network is applied to tune the gains of the PD controller. The error of the Euler angles is employed as an input of the neural network. The closed-loop system with different characteristics is simulated and compared with the variable-structure controller. The results show the superior performance of the proposed controller. [ABSTRACT FROM AUTHOR]

Traffic flow forecasting, especially the short-term case, is an important topic in intelligent transportation systems (ITS). This paper researches network-scale modeling and forecasting of short-term traffic flows. First, the concepts of single-link and multilink models of traffic flow forecasting are proposed. Secondly, four prediction models are constructed by combining the two models with single-task learning (STL) and multitask learning (MTL). The combination of the multilink model and multitask learning not only improves the experimental efficiency but also improves the prediction accuracy. Moreover, a new multilink, single-task approach that combines graphical lasso (GL) with neural network (NN) is proposed. GL provides a general methodology for solving problems involving lots of variables. Using L1 regularization, GL builds a sparse graphical model, making use of the sparse inverse covariance matrix. Gaussian process regression (GPR) is a classic regression algorithm in Bayesian machine learning. Although there is wide research on GPR, there are few applications of GPR in traffic flow forecasting. In this paper, GPR is applied to traffic flow forecasting, and its potential is shown. Through sufficient experiments, all of the proposed approaches are compared, and an overall assessment is made. [ABSTRACT FROM AUTHOR]

This paper introduces a new fault identification method that uses pseudomodal energies to train probabilistic neural networks (PNNs). The proposed procedure is tested on a population of 20 cylindrical shells and its performance is compared to the procedure which uses modal properties to train probabilistic neural networks. The PNNs trained using pseudomodal energies provide better classification of faults than the PNNs trained using the conventional modal properties. [ABSTRACT FROM AUTHOR]

In this paper, a new type of multilayer feedforward neural network has been proposed based on inspiration from the Preisach model, which has been called the Preisach neural network (Preisach-NN). It is comprised of input, output, and two hidden layers. The input and output layers contain linear neurons, whereas the first hidden layer incorporates neurons called stop neurons, whose activation function represents a stop operator. The second hidden layer includes sigmoidal neurons. The subgradient optimization method with space dilatation has been adopted for training of the Preisach-NN as a nonsmooth problem. Although the proposed Preisach-NN could be mathematically identical to the Preisach model, tuning of the Preisach-NN is easier and also more general than that of the model. To assess their capability, Preisach-NNs are used to model two different types of hysteretic behaviors of Masing and non-Masing problems. The results presented and discussed in this paper show that the neural networks have been capable of learning the material behaviors successfully and with high precision. [ABSTRACT FROM AUTHOR]

This paper examines a linked simulation-optimization procedure based on the combined application of an artificial neural network (ANN) and genetic algorithm (GA) with the aim of developing an efficient model for the multiobjective management of groundwater lenses in small islands. The simulation-optimization methodology is applied to a real aquifer in Kish Island of the Persian Gulf to determine the optimal groundwater-extraction while protecting the freshwater lens from seawater intrusion. The initial simulations are based on the application of SUTRA, a variable-density groundwater numerical model. The numerical model parameters are calibrated through automated parameter estimation. To make the optimization process computationally feasible, the numerical model is subsequently replaced by a trained ANN model as an approximate simulator. Even with a moderate number of input data sets based on the numerical simulations, the ANN metamodel can be efficiently trained. The ANN model is subsequently linked with GA to identify the nondominated or Pareto-optimal solutions. To provide flexibility in the implementation of the management plan, the model is built upon optimizing extraction from a number of zones instead of point-well locations. Two issues are of particular interest to the research reported in this paper are: (1) how the general idea of minimizing seawater intrusion can be effectively represented by objective functions within the framework of the simulation-optimization paradigm, and (2) the implications of applying the methodology to a real-world small-island groundwater lens. Four different models have been compared within the framework of multiobjective optimization, including (1) minimization of maximum salinity at observation wells, (2) minimization of the root mean square (RMS) change in concentrations over the planning period, (3) minimization of the arithmetic mean, and (4) minimization of the trimmed arithmetic mean of concentration in the observation wells. The latter model can provide a more effective framework to incorporate the general objective of minimizing seawater intrusion. This paper shows that integration of the latest innovative tools can provide the ability to solve complex real-world optimization problems in an effective way. [ABSTRACT FROM AUTHOR]

The performance of a fuzzy expert system (FES) is significantly affected by the accuracy of its knowledge base parameters (membership functions and rule bases). The main contribution of this paper is in presenting a methodology to integrate an FES with adaptation/optimization techniques and applying the data-based adaptive learning concept to increase the accuracy of an FES developed for contractor default prediction for surety bonding. In addition, this paper investigates two optimization approaches (genetic algorithms and neural network back-propagation) for adaptation of fuzzy membership function (MBF) and rules' degree of support (DoS) to determine the most suitable technique to adapt the FES. The optimized FES, called SuretyQualification, was validated using 30 hypothetical contractor default prediction cases, and the highest accuracy of the system (adapted using neural networks) was found to be 91.83%. Another contribution of this paper is the development of a software tool called SuretyQualification that provides a comprehensive and systematic evaluation process to evaluate a contractor and their risk of default on a project. The presented optimization approaches address FES context adaptation using any changing information conveyed by the input-output data and provide a methodology for continuous adaptation of the FES parameters, using practical cases to adjust the FES according to any contexts changes. [ABSTRACT FROM AUTHOR]

In this paper, the use of a hybrid evolutionary optimization algorithm is proposed for global optimization of pavement structural parameters through inverse modeling. Shuffled complex evolution (SCE) is a population-based stochastic optimization technique combining the competitive complex evolution with the controlled random search, the implicit clustering, and the complex shuffling. Back-calculation of pavement layer moduli is an ill-posed inverse engineering problem, which involves searching for the optimal combination of pavement layer stiffness solutions in an unsmooth, multimodal, complex search space. SCE is especially considered a robust and efficient approach for global optimization of multimodal functions. A desirable characteristic of the SCE algorithm is that it uses information about the nature of the response surface, extracted using the deterministic Simplex geometric shape, to direct the search into regions with higher posterior probability. The hybrid back-calculation system described in this paper combines the robustness of the SCE in global optimization with the computational efficiency of neural networks and advanced pavement system characterization offered by employing finite-element models. This is the first time the SCE approach is applied to real-time nondestructive evaluation of pavement systems required in the routine maintenance and rehabilitation activities for sustainable transportation infrastructure. [ABSTRACT FROM AUTHOR]

This paper presents an enhanced design methodology for optimal mixture proportion of concrete composition with respect to accuracy in the case of using prediction models based on a limited database. In proposed methodology, the search space is constrained as the domain defined by a limited database instead of constructing the database covering the region represented by the possible ranges of all variables in the input space. A model for defining the search space which is expressed by the effective region in this paper and evaluating whether a mix proportion is effective is added to the optimization process, yielding highly reliable results. To demonstrate the proposed methodology, a genetic algorithm, an artificial neural network, and a convex hull were adopted as an optimum technique, a prediction model for material properties, and an evaluation model for the effective region, respectively. And then, it was applied to an optimization problem wherein the minimum cost should be obtained under a given strength requirement. Experimental test results show that the mix proportion obtained from the proposed methodology considering the regional characteristics of the database is found to be more accurate and feasible than that obtained from a general optimum technique that does not consider this aspect. [ABSTRACT FROM AUTHOR]

This paper describes the use of inductive models developed using two artificial intelligence (AI)-based techniques for fecal coliform prediction and classification in surface waters. The two AI techniques used include artificial neural networks (ANNs) and a fixed functional set genetic algorithm (FFSGA) approach for function approximation. While ANNs have previously been used successfully for modeling water quality constituents, FFSGA is a relatively new technique of inductive model development. This paper will evaluate the efficacy of this technique for modeling indicator organism concentrations. In scenarios where process-based models cannot be developed and/or are not feasible, efficient and effective inductive models may be more suitable to provide quick and reasonably accurate predictions of indicator organism concentrations and associated water quality violations. The relative performance of AI-based inductive models is compared with conventional regression models. When raw data are used in the development of the inductive models described in this paper, the AI models slightly outperform the traditional regression models. However, when log transformed data are used, all inductive models show comparable performance. While the work validates the strength of simple regression models, it also validated FFSGA to be an effective technique that competes well with other state-of-the-art and complex techniques such as ANNs. FFSGA comes with the added advantage of resulting in a simple, easy to use, and compact functional form of the model sought. This work adds to the limited amount of research on the use of data-driven modeling methods for indicator organisms. [ABSTRACT FROM AUTHOR]

The height difference between the ellipsoidal height h and the orthometric height H is called undulation N. The key issue in transforming the global positioning system (GPS)-derived ellipsoidal height to the orthometric height is to determine the undulation value accurately. If the undulation N for a point whose position is determined by a GPS receiver can be estimated in the field, then the GPS-derived three-dimensional geocentric coordinate in WGS-84 can be transformed into a local coordinate system and the orthometric height in real-time. In this paper, algorithms of applying a back-propagation artificial neural network (BP ANN) to develop a regional grid-based geoid model using GPS data (e.g., ellipsoidal height) and geodetic leveling data (e.g., orthometric height) are proposed. In brief, the proposed algorithms include the following steps: (1) establish the functional relationship between the point’s plane coordinates and its undulation using the BP ANN according to the measured GPS data and leveling data; (2) develop a regional grid-based geoid model using the imaginary grid plane coordinates with a fixed grid interval and the trained BP ANN; (3) develop an undulation interpolation algorithm to estimate a specific point’s undulation using the generated grid-based geoid model; and (4) estimate the point’s undulation in the field and transform the GPS ellipsoidal height into the orthometric height in real-time. Three data sets from the Taiwan region are used to test the proposed algorithms. The test results show that the undulation interpolation estimation accuracy using the generated grid-based geoid is in the order of 2–4 cm. The proposed algorithms and the detailed test results are presented in this paper. [ABSTRACT FROM AUTHOR]

Over the years, many methods have been developed to predict the concrete strength. In recent years, artificial neural networks (ANNs) have been applied to many civil engineering problems with some degree of success. In the present paper, ANN is used as an attempt to obtain more accurate concrete strength prediction based on parameters like concrete mix design, size and shape of specimen, curing technique and period, environmental conditions, etc. A total of 864 concrete specimens were cast for compressive strength measurement and verification through the ANN model. The back propagation-learning algorithm is employed to train the network for extracting knowledge from training examples. The predicted strengths found by employing ANN are compared with the actual values. The results indicate that ANN is a useful technique for predicting the concrete strength. Further, an effort to build an expert system for the problem is described in this paper. To overcome the bottleneck of intricate knowledge acquisition, an expert system is used as a mechanism to transfer engineering experience into usable knowledge through rule-based knowledge representation techniques. [ABSTRACT FROM AUTHOR]

This paper reports on the development of a new approach for simulating the thermal behavior of buildings that overcome the limitations of conventional heat-transfer simulation methods such as the finite difference method and the finite element method. The proposed technique uses a coarse-grain approach to model development whereby each element represents a complete building component such as a wall, internal space, or floor. The thermal behavior of each coarse-grain element is captured using empirical modeling techniques such as artificial neural networks (ANNs). The main advantages of the approach compared to conventional simulation methods are (1) simplified model construction for the end-user; (2) simplified model reconfiguration; (3) significantly faster simulation runs (orders of magnitude faster for two- and three-dimensional models); and (4) potentially more accurate results. The paper demonstrates the viability of the approach through a number of experiments with a model of a composite wall. The approach is shown to be able to sustain highly accurate long-term simulation runs, if the coarse-grain modeling elements are implemented as ANNs. In contrast, an implementation of the coarse-grain elements using a linear model is shown to function inaccurately and erratically. The paper concludes with an identification of on-going work and future areas for development of the technique. [ABSTRACT FROM AUTHOR]

In this paper, functional networks (FN) proposed by Castillo as an alternative to neural networks are discussed. Unlike neural networks, the functions are learned instead of weights. In general, topology is selected based on data, domain knowledge (properties of the function such as associativity, commutativity, and invariance), or a combination of the two. The object of this paper is to show the application of some functional network architectures to model and predict the behavior of structural systems which are otherwise modeled in terms of differential or difference equations or in terms of neural networks. In this paper, four examples in structural engineering and one example in mathematics are discussed. The results obtained by functional networks are compared with those obtained by neural networks for the first four examples, and it is shown that functional networks are more efficient and powerful and take much less computer time as compared to predictions by conventional neural networks such as the back-propagation network. [ABSTRACT FROM AUTHOR]

In India, presently 85 existing thermal power plants (TPPs) produce 118 million metric tons of fly ash per year, which is projected to exceed 170 and 440 million metric tons/year by Years 2012 and 2030, respectively. This huge quantity of fly ash not only poses the problems of environmental concerns but occupies large areas of land for its dumping, which needs urgent and appropriate measures for its safe disposal and gainful utilization on sustainable basis. Based on the detailed study on the bulk utilization of pond ash in agriculture and forestry sectors under different agroclimatic conditions and soil types for last 2 decades, it has been found that pond ash has a very good potential for the utilization on bulk scale as liming agent, soil conditioner, source of essential plant nutrients, and also for boosting the growth and yields of a variety of crops and growth performance of plant species. Some field-scale studies carried out especially in the wastelands of State Agriculture Research Farm and farmers' fields are discussed in the present paper. The influence of critical parameters on the yield of wheat crops cultivated in different soil types and agroclimatic conditions is discussed. Further, an attempt has been made to develop a threelayer feed-forward artificial neural network (ANN) model, which is inherently trained using error back-propagation algorithm. The results evince that the predictions from the ANN model are in good qualitative and quantitative agreement with the field observations, thereby validating the applicability and accuracy of the developed ANN model. [ABSTRACT FROM AUTHOR]

Many of the empirical formulas used for the prediction of the expected scour depth at piers are excessively conservative, providing substantial overestimations. On the other hand, the recently proposed neural networks methods generally issue accurate predictions but also high percentages of underpredictions, due to the use of a symmetric error function for their parameterization. A novel error function is proposed in this paper for optimizing neural networks, giving more weight to underestimation than to overestimation discrepancies, in order to obtain safer design predictions. The performances of the proposed model on independent field records are compared with those of a conventionally trained neural network and with those of a set of widely used formulas. The asymmetric error function (that might be applied to parameterize any other model or equation, as a proficient alternative to least-square errors or envelope curves) allows researchers to obtain predictions closer to the measurements than those issued by traditional formulas, substantially reducing the extent of unnecessary overdesign and at the same time the percentage of severe underestimations is comparable with those of the safest formulas. [ABSTRACT FROM AUTHOR]

In recognition of the importance of historical knowledge in decision making, case based reasoning (CBR) is utilized as a form of an expert system to tackle construction management issues such as quantity takeoff in the proposal development phase of a project. It builds on a proposition that past projects similar to the new one would suggest a reasonable range of craft quantities. This paper finds that when measuring the similarity between the new project and historical projects, traditional similarity measure methods fail to consider the nonlinearity and muticollinearity embedded in the problem, as well as differences across crafts. An innovative similarity measurement algorithm was therefore proposed to tackle the above issues with a carefully designed orthogonalization process and Sobol's total sensitivity analysis. The application of the proposed algorithm to the craft quantity takeoff of a power plant project was introduced, demonstrating a better result compared with traditional methods. It is likely that the proposed algorithm will advance current CBR practices in construction management. [ABSTRACT FROM AUTHOR]

In this paper, an adaptive neural network controller is proposed for a nonaffine-nonlinear aircraft cabin pressure system with unknown parameters. A multilayer neural network is used to represent the controller structure. The ultimate boundedness of the closed-loop system is guaranteed through a Lyapunov stability analysis by introducing a suitably driven adaptive rule. The effectiveness of the proposed adaptive controller is illustrated by considering an aircraft cabin pressure system, and the simulation results verify the merits of the proposed controller. [ABSTRACT FROM AUTHOR]

Missing values in in situ monitoring data is a problem often encountered in hydrologic research and applications. Values in a data set may be missing because of sensor error or failure of data recording devices. Whereas various imputation techniques have focused on hydrometeorological data, very few studies have investigated gap-filling methods for soil moisture data. This paper aims to fill that gap by investigating well-established statistical and data-driven methods for infilling missing values in a high resolution, soil moisture time series. Since 2006, the authors collected hourly soil moisture data in the Hamilton-Halton Watershed, Southern Ontario, Canada at four research sites. Each site contained nine stations with time domain reflectometry (TDR) soil sensors at six soil depths. From these distributed data sets, the authors removed values randomly () and systematically () from the data to evaluate the effectiveness of the monthly average replacement (MAR), soil layer relative difference (SLRD), linear and cubic interpolation, artificial neural networks (ANN), and evolutionary polynomial regression (EPR) infilling methods. When values were randomly removed, interpolation, ANN, and EPR were able to infill the missing values with similar efficiency, whereas MAR and SLRD were the least effective methods. Similarly, when large systematic gaps were present in the data, interpolation and ANN were the most effective methods of infilling, respectively. However, the effectiveness of both infilling methods is limited as serial gaps become larger than 72-100 h. [ABSTRACT FROM AUTHOR]

The general swelling model has recently been updated in Israel by applying the Excel-solver command (ESC) analysis to new local test results from 897 undisturbed specimens. In this analysis, the goodness-of-fit statistics obtained classify the category of their associated regression only as fair. Thus, it seems necessary to explore the possibility of enhancing the outputs of this regression analysis by applying the artificial neural networks (ANN) methodology to the same 897 undisturbed specimens. However, it is shown that the use of the ANN outputs should be accompanied by an additional check to ensure that they follow the expected physical swelling behavior, as characterized by the index properties of the soil. The ANN methodology applied in this paper is similar to previous studies in geotechnical engineering. Different models were tested using the same database (i.e., the same 897 undisturbed specimens). The statistical fit of the ANN models were clearly found to be superior to the ESC models. However, in the sense of the required physical behavior, as characterized by the index properties of the soil, the ANN models did not predict swelling values as well as ESC models did, in particular values ranging near (or outside) the data set boundaries. Thus, the former ESC models still remain preferable. [ABSTRACT FROM AUTHOR]

In this paper, a novel adaptive control approach, named adaptive Chebyshev retrofit control (ACRC), retrofitting an existing baseline controller with an adaptive Chebyshev function approximator is presented. The approximator is composed of a linear combination of a parameter and a basis function. Instead of using neural networks as a function approximator, the new approach utilizes a Chebyshev polynomial as a basis function for function approximation, and a parameter update law is derived via a Lyapunov-like analysis method. The benefits of the proposed method are twofold. First, the computational time is approximately 1.7 times faster than that of the method using the neural network. Second, the implementation is very efficient, because the structure of the approximator is significantly simpler in comparison with those of neural network approaches. Because the complexity of the software is the major contributing factor to software reliability, the high complexity of the implementation of a control algorithm that adopts neural networks could lead to a reduction in software reliability. Therefore, the new adaptive control method is valuable in terms of the improvement in software reliability. In particular, it is important in the field of aerospace control, which requires exceptional reliability for flight control software. Moreover, the short computational time in comparison with neural network approaches is very crucial for small unmanned aerial vehicles that have restricted on-board hardware performance. From simulation results, it is found that the performance of the proposed method in several responses is on par with that of the neural network method in the presence of varying flight conditions. Considering the computation time and simplicity of the proposed method, the authors conclude that the proposed approach is very effective, particularly relative to the neural network method. [ABSTRACT FROM AUTHOR]

Complex analysis and design of structures, especially landmark structures such as long-span bridges, have been conducted by many engineers and researchers. Currently, it is possible to collect more and precise monitoring data as well as to develop complex three-dimensional (3D) FEM models. These models, which can be calibrated using structural health-monitoring (SHM) data, can be used for the estimation of component and system reliability of bridges. However, the uncertainties related to the data, analysis, and nonstationary nature of the structural behavior need to be better incorporated by using a set of models that are continuously updated with monitoring data. This set of models constitutes a family as a result of the approach by which the models are obtained and the relationships among them. The objective of this paper is to explore the impact of uncertainty in predicting the system reliability obtained by a one-time, initially calibrated FEM model as well as by a family of FEM models continuously calibrated with monitoring data. To explore the uncertainty effects, a laboratory structure that has a combined system configuration with main and secondary elements is monitored. The monitoring data are employed for the FEM model calibration by using artificial neural networks (ANNs) to obtain parent (calibrated) FEM models from which a set of offspring FEM models is generated to incorporate the uncertainties. It is shown that the use of parent-offspring FEM models becomes important especially when critical parameters that have an impact on the model responses cannot be precisely defined. Finally, it is shown in a comparative fashion that the prediction of reliability using a family of FEM models and a single model can be quite different because the family of models provides a more realistic estimate of the structural responses and probability of failure. [ABSTRACT FROM AUTHOR]

In this paper, models for short-term prediction of influent flow rate in a wastewater-treatment plant are discussed. The prediction horizon of the model is up to 180 min. The influent flow rate, rainfall rate, and radar reflectivity data are used to build the prediction model by different data-mining algorithms. The multilayer perceptron neural network algorithm has been selected to build the prediction models for different time horizons. The computational results show that the prediction model performs well for horizons up to 150 min. Both the peak values and the trends are accurately predicted by the model. There is a small lag between the predicted and observed influent flow rate for horizons exceeding 30 min. The lag becomes larger with the increase of the prediction horizon. [ABSTRACT FROM AUTHOR]

Delamination of the concrete cover above the upper reinforcing bars is a common problem in concrete bridge decks. Acoustic nondestructive evaluation is widely used to detect such delamination because of its low cost, fast speed, and ease of implementation. The accuracy of traditional acoustic approaches is dependent on the level of ambient noise, and the detection process is highly subjective. An automatic impact-based delamination detection (AIDD) system is described in this paper. In this system, the traffic noise is eliminated by a modified version of independent component analysis. Mel-frequency cepstral coefficients are then used as features for detection to eliminate subjectivity. The delamination detection is performed by a radial basis function neural network. The AIDD system was developed using mixed-language programming in MATLAB, LabVIEW, and C++. The performance of the system was evaluated using data from two bridges, and the results were satisfactory. [ABSTRACT FROM AUTHOR]

This paper presents an automatic landing control design using adaptive, integrated guidance and control (IGC) logic. The proposed IGC design uses a combination of an adaptive output feedback inversion and backstepping techniques. The problem is formulated as an adaptive output feedback control problem for a line-of-sight-based chasing flight configuration. The design objective is to regulate the relative distance between the aircraft and the moving reference point on a landing pattern and two bearing angles maintaining turn coordination. Adaptive neural networks are trained online with available measurements to compensate for inversion error as a result of unmodeled dynamics and modeling errors of the aircraft in the design process. In addition, a reference command trajectory for the automatic landing control is designed in a way that the aircraft follows the landing pattern regardless of its initial position. The automatic landing system using IGC logic is evaluated using a sophisticated six-degrees-of-freedom nonlinear simulation program with the approach and landing scenario. [ABSTRACT FROM AUTHOR]

Over the years of the study (1999–2004) presented in this paper, the City of Edmonton, Canada’s Drainage and Maintenance Department has experienced an annual increase of about 12% in the installation of water and sewer services for residential facilities. According to the current estimating procedure, a discrepancy of up to 60% exists between the estimated and actual costs of these projects. A detailed analysis of all activities involved in the installation of the water and sewer services has been carried out and is presented in this paper. The proposed methodology, which is based upon the analysis of past data obtained from the City of Edmonton’s drainage division for the period of 1999–2004, is also presented. The methodology has been incorporated into a computer module, which integrates the concept of artificial neural network (ANN) with the current estimating system used by the City of Edmonton. The following research includes a description of the algorithm used in ANN, as well as an assessment of past data obtained from the city record for over 800 jobs (cases) performed over the period of the study. [ABSTRACT FROM AUTHOR]

This paper describes a novel hybrid intelligent system approach to inversion of nondestructive pavement deflection data and back-calculation of nonlinear stress-dependent pavement layer moduli. Particle swarm optimization (PSO), a population-based stochastic optimization technique inspired by social behavior of bird flocking or fish schooling, is fast emerging as an innovative and powerful computational metaphor for solving complex problems in design, optimization, control, management, business, and finance. Back-calculation of pavement layer moduli is an ill-posed inverse engineering problem which involves searching for the optimal combination of pavement layer stiffness solutions in an unsmooth, multimodal, complex search space. PSO is especially considered a robust and efficient approach for global optimization of multimodal functions. The hybrid back-calculation system described in this paper integrates finite element modeling, neural networks, and PSO in an efficient manner to mitigate the limitations and take advantages of the strengths to produce a system that is more effective and powerful than those which could be built with single technique. This is the first time the PSO approach is applied to real-time nondestructive evaluation of pavement systems. [ABSTRACT FROM AUTHOR]

This paper presents a coupled approach using an artificial neural network (ANN) and the finite difference method (FDM) that has been developed to predict the distribution of axial load along fully grouted standard cable bolts in the field using laboratory pullout test data. A back-propagation training algorithm was used in ANN to determine axial loads in the cables tested in the laboratory. The ANN component of the computational model was trained using two different types of data sets. At first, the ANN was trained to predict the axial loads in a series of short cables grouted with Portland cement at a specific water-to-cement ratio and subjected to different radial confining stiffness values. Next, the ANN model was trained for an expanded case to include the influence of lateral confining stress on the distribution of axial load in the cable reinforcement. Finally, the ANN model was implemented into a widely used, FDM-based geotechnical software (FLAC). The accuracy of the ANN–FDM model is demonstrated in this paper against measured data from laboratory and field tests. The analysis approach introduced in this study is a valuable computational tool that can be used to determine the axial load distribution in long standard cable bolts, which are commonly installed to stabilize rock masses in various geotechnical, transportation, and mining applications. [ABSTRACT FROM AUTHOR]

This paper deals with the challenging problem of hydrological interpretation of the internal functioning of artificial neural networks (ANNs) by extracting knowledge from their solutions. The neural network used in this study is based on the structure of the radial-basis-function neural network (RBFNN), which is considered as an alternative to the multilayer perceptron for solving complex modeling problems. This network consists of input, hidden, and output layers. The network is trained using the daily data of two catchments having different characteristics and from two different regions in the world. The present day and antecedent observed discharges are used as inputs to the network to forecast the flow one day ahead. A range of quantitative and qualitative techniques are used for hydrological interpretation of the internal functioning by examining the responses of the hidden layer nodes. The results of the study show that a single hidden layered RBFNN is an effective tool to forecast the daily flows and that the activation of the hidden layer nodes are far from arbitrary, but appear to represent flow components of the predicted hydrograph. The results of the study confirm that the three nodes in the hidden layer of this model effectively divide the input data space in such a way that the contribution from each node dominates in one of the flow domains—low, medium, or high—and form, in a crude manner, the base flow, interflow and surface runoff components of the hydrograph. [ABSTRACT FROM AUTHOR]

Bridge risk assessment often serves as the basis for bridge maintenance priority ranking and optimization and conducted periodically for the purpose of safety. This paper presents an application of artificial neural networks in bridge risk assessment, in which back-propagation neural networks are developed to model bridge risk score and risk categories. The study investigated and utilized 506 bridge maintenance projects to develop the models. It is shown that neural networks have a very strong capability of modeling and classifying bridge risks. The average accuracies for risk score and risk categories are both over 96%. A comparative study is conducted with an alternative methodology using multiple regression techniques. The results revealed that neural networks achieved much better performances than regression analysis models. In addition an integrated forecasting approach was utilized to combine neural networks and regression analysis to generate hybrid models, which produced better accuracies than any of the individually developed models. [ABSTRACT FROM AUTHOR]

The impact of topography on design wind speed is addressed in current building codes and design standards by providing “speed-up” ratios for limited cases of terrain geometries. This paper proposes a combined numerical–neural network (NN) approach to provide speed-up ratios for a wide range of topographic features such as single and multiple hills, escarpments, and valleys. In this approach learning data required by the NN is generated via a detailed numerical approach based on computational fluid dynamics (CFD). Use of the developed model only requires simple geometrical input such as slope, height, and ground roughness while producing results of comparable accuracy to complex numerical evaluations. This combined CFD-NN approach not only produces data for new cases but also conveys the results of complex CFD simulations to the engineering profession (end user). Results compare well with an independent set of experimental data demonstrating the feasibility of the CFD-NN approach to generate data to apply wind design load provisions to buildings with upstream complex terrain. [ABSTRACT FROM AUTHOR]

This paper describes investigations into a development of a new application of neural networks (NN) for prediction of pipeline failure. Results show higher correlations with recorded data when compared with the two existing statistical models. The shifted time power model gives results in total number of failures and the shifted time exponential model gives results in number of failures per year. The database was large but neither complete and nor fully accurate. Factors influencing pipeline deterioration were missing from the database. Using the NN technique on this database produced models of pipeline failure, in terms of failures/km/year, that more closely matched the number of failures of a particular asset recorded for the period. [ABSTRACT FROM AUTHOR]

This paper investigates the influence of different hydrological input data on the management simulation of a water resources system (WRS). Three complete simulations were carried out using synthetic inflow series generated with several stochastic models: an autoregressive moving average (ARMA) model, the Lane condensed temporal disaggregation model, and a nonlinear model based on a multilayer perceptron artificial neural network (MLP-ANN) with a random component embedded. The validation of the stochastic models was performed using comparisons of relevant drought statistics from synthetic series with those from the historical records. Since the analyzed WRS includes five inflow sites, multivariate models were applied. The MLP-ANN model showed the best performance. The management simulations of the WRS were executed with the decision support system AQUATOOL under a probabilistic approach. This approach gives probabilities of demand failures of the WRS, which were used to evaluate the influence of the three applied stochastic models on the simulation results. Significant differences were found. [ABSTRACT FROM AUTHOR]

Based on the basic formulation developed in a companion paper, the writers now present the application of an artificial neural network approach to designing streamlined network models to simulate and identify the nonlinear dynamic response of single-degree-of-freedom oscillators using the restoring force-state mapping interpretation. The neural networks which use sigmoidal activation functions are shown to be highly robust in modeling a wide variety of commonly observed nonlinear structural dynamic response behaviors. By streamlining the networks, individual network model parameters take on physically or geometrically interpretable meaning, and hence, the network initialization can be achieved through an engineered approach rather than through less physically meaningful numerical initialization schemes. Although not proven in general, examples show that by starting with a more meaningful initial design, identification convergence is improved, and the final identified model parameters are seen to have a more physical meaning. A set of model architecture prototypes is developed to capture commonly observed nonlinear response behaviors. [ABSTRACT FROM AUTHOR]

This paper deals with an application of neural networks for computation of fundamental natural periods of buildings with load-bearing walls. The analysis is based on long-term tests performed on actual structures. The identification problem is formulated as the relation between structural and soil basement parameters, and the fundamental period of building. The principal component analysis for compression of input data is also used. Backpropagation neural networks are applied in the analysis. Results of neural network identification of natural periods are compared with data from experiments. The application of the proposed neural networks enables us to identify the natural periods of the buildings with quite satisfactory accuracy for engineering practice. The compression of the input data to principal components by principal component analysis makes it possible to design much smaller neural networks than those without data compression with no greater increase of the neural approximation errors. It appears that this technique would also be very useful in damage detection and health monitoring of structures. [ABSTRACT FROM AUTHOR]

Fly ash and slag concrete (FSC) is a highly complex material whose behavior is difficult to model. This paper describes a method of modeling slump of FSC using artificial neural networks. The slump is a function of the content of all concrete ingredients, including cement, fly ash, blast furnace slag, water, superplasticizer, and coarse and fine aggregate. The model built was examined with response trace plots to explore the slump behavior of FSC. This study led to the conclusion that response trace plots can be used to explore the complex nonlinear relationship between concrete components and concrete slump. [ABSTRACT FROM AUTHOR]

Short-term traffic flow prediction has long been regarded as a critical concern for intelligent transportation systems. On the basis of many existing prediction models, each having good performance only in a particular period, an improved approach is to combine these single predictors together for prediction in a span of periods. In this paper, a neural network model is introduced that combines the prediction from single neural network predictors according to an adaptive and heuristic credit assignment algorithm based on the theory of conditional probability and Bayes’ rule. Two single predictors, i.e., the back propagation and the radial basis function neural networks are designed and combined linearly into a Bayesian combined neural network model. The credit value for each predictor in the combined model is calculated according to the proposed credit assignment algorithm and largely depends on the accumulative prediction performance of these predictors during the previous prediction intervals. For experimental test, two data sets comprising traffic flow rates in 15-min time intervals have been collected from Singapore’s Ayer Rajah Expressway. One data set is used to train the two single neural networks and the other to test and compare the performances between the combined and singular models. Three indices, i.e., the mean absolute percentage error, the variance of absolute percentage error, and the probability of percentage error, are employed to compare the forecasting performance. It is found that most of the time, the combined model outperforms the singular predictors. More importantly, for a given time period, it is the role of this newly proposed model to track the predictors’ performance online, so as to always select and combine the best-performing predictors for prediction. [ABSTRACT FROM AUTHOR]

The Food and Agriculture Organization of the United Nations (FAO) has proposed using the Penman–Monteith (FAO-56 PM) method as the standard method for estimating reference evapotranspiration (ET0), and for evaluating other methods. The basic obstacle to widely using this method is the numerous required data that are not available at many weather stations. The maximum and minimum air temperatures constitute a set of minimum data necessary for the estimation of ET0. The basic goal of the paper is to examine whether it is possible to attain the reliable estimation of ET0 only on the basis of the temperature data. This goal was reached by the evaluation of the reliability of four temperature-based approaches [radial basis function (RBF) network, Thornthwaite, Hargreaves, and reduced set Penman–Monteith methods] as compared to the FAO-56 PM method. The seven weather stations selected for this study are located in Serbia (Southeast Europe). The Thornthwaite, Hargreaves, and reduced set Penman–Monteith methods mostly underestimated or overestimated ET0 obtained by the FAO-56 PM method. In this study, methods were calibrated using the standard FAO-56 PM method. However, the RBF network better predicted FAO-56 PM ET0 than calibrated temperature-based methods at most locations. It gives reliable results in all locations and it has proven to be the most adjustable to the local climatic conditions. These results are of significant practical use because the adaptive temperature-based RBF network can be used when relative humidity, radiation, and wind speed data are not available. [ABSTRACT FROM AUTHOR]

Engineering and design professionals constitute a major driving force for a successful project undertaking. Although the industry has been active in addressing the performance of construction labor and methods to estimate or predict such performance, relatively fewer efforts have been conducted for the engineering profession. In an attempt to fill out this gap, the paper presents a study to utilize neurofuzzy intelligent systems for predicting the engineering performance in a construction project. First, neurofuzzy systems are introduced as integrated schemes of artificial neural networks and fuzzy control systems. The use of these neurofuzzy intelligent systems, particularly fuzzy neural networks, in predicting engineering performance is then demonstrated in the industrial construction sector. The development of the system is based on actual project data that was collected through questionnaire surveys. Statistical variable reduction techniques are further employed to develop linear regression models of the same engineering performance prediction scheme, and results are being compared between both techniques. [ABSTRACT FROM AUTHOR]

A new method of estimating flutter derivatives using artificial neural networks is proposed. Unlike other computational fluid dynamics based numerical analyses, the proposed method estimates flutter derivatives utilizing previously measured experimental data. One of the advantages of the neural networks approach is that they can approximate a function of many dimensions. An efficient method has been developed to quantify the geometry of deck sections for neural network input. The output of the neural network is flutter derivatives. The flutter derivatives estimation network, which has been trained by the proposed methodology, is tested both for training sets and novel testing sets. The network shows reasonable performance for the novel sets, as well as outstanding performance for the training sets. Two variations of the proposed network are also presented, along with their estimation capability. The paper shows the potential of applying neural networks to wind force approximations. [ABSTRACT FROM AUTHOR]

Seismic early warning has been very important and has become feasible in Taiwan. Perhaps because of the lack of quick and reliable estimations of the induced structural response, however, the triggering criteria of almost all of the existing earthquake protection or early warning systems in the world are merely based on the collected or estimated data of the ground motion, without any information regarding the structural response. This paper presents a methodology of generating quick seismic response estimations of a prestressed concrete (PC) bridge using artificial neural networks (ANNs), which may be incorporated in a seismic early warning system for the bridge. In the methodology ANNs were applied to model the critical structural response of a PC bridge subjected to earthquake excitation of various magnitudes along various directions. The objective was to implement a well-trained network that is capable of providing a quick prediction for the critical response of the target bridge. The well-known multilayer perception (MLP) networks with back propagation algorithm were employed. A simple augmented form of MLP that can be quantitatively determined was proposed. These networks were trained and tested based on the analytical data obtained from the nonlinear dynamic finite fiber element analyses of the target PC bridge. The augmented MLPs were found to be much more efficient than the MLPs in modeling the critical bending moments of the piers and girder of the PC bridge. [ABSTRACT FROM AUTHOR]

In recent years, artificial neural networks (ANNs) have been applied to many geotechnical engineering problems with some degree of success. In the majority of these applications, data division is carried out on an arbitrary basis. However, the way the data are divided can have a significant effect on model performance. In this paper, the issue of data division and its impact on ANN model performance is investigated for a case study of predicting the settlement of shallow foundations on granular soils. Four data division methods are investigated: (1) random data division; (2) data division to ensure statistical consistency of the subsets needed for ANN model development; (3) data division using self-organizing maps (SOMs); and (4) a new data division method using fuzzy clustering. The results indicate that the statistical properties of the data in the training, testing, and validation sets need to be taken into account to ensure that optimal model performance is achieved. It is also apparent from the results that the SOM and fuzzy clustering methods are suitable approaches for data division. [ABSTRACT FROM AUTHOR]

The neuro-fuzzy refers to the recent technology that couples the traditional fuzzy logic developments with neural nets training capabilities to compose the fuzzy logic’s knowledge base and fuzzy sets’ parameters optimally. This paper discusses the calibration methodology of a neuro-fuzzy logic for modeling the route choice behavior. The logic accounts for the various factors of potential effect on the route choice utility perceived by the traveler. The structure of the fuzzy control stages, the calibration of the membership functions, and the composition of the knowledge base are discussed in detail. Logic training is based on data extracted from a factorial experimental design model. The results of the fuzzy logic model are utilized for in-depth analyses of the travelers’ perceptions of the route utility in response to the various traffic states. [ABSTRACT FROM AUTHOR]

Because the complexity of sulfur capture and release during solid-fuel combustion in a circulating fluidized bed combustors (CFBC), especially in the oxygen-enriched combustion, has not been sufficiently recognized, the development of a simple model, which can correctly predict the SO2 emissions from such units over a wide range of operating conditions is of practical significance. The artificial neural network (ANN) approach is proposed in this paper, which may overcome the shortcomings of the experimental procedures and the programmed computing approach. The Ca:S molar ratio, oxygen concentration in inlet gas, excess oxygen, average riser temperature, mean diameter of the coal particles, average gas velocity in the riser, flue gas recycle ratio, and inlet gas pressure are taken into account by the model as the input parameters. The [8-3-7-1] ANN model with hyperbolic tangent sigmoid activation function was successfully applied to calculate the SO2 emissions from coal combustion in several CFB boilers operating under both air-fired and oxygen-enriched conditions. [ABSTRACT FROM AUTHOR]

The design of pile foundations requires good estimation of the pile load-carrying capacity and settlement. Design for bearing capacity and design for settlement have been traditionally carried out separately. However, soil resistance and settlement are influenced by each other, and the design of pile foundations should thus consider the bearing capacity and settlement inseparably. This requires the full load-settlement response of piles to be well predicted. However, it is well known that the actual load-settlement response of pile foundations can be obtained only by load tests carried out in situ, which are expensive and time-consuming. In this paper, recurrent neural networks (RNNs) were used to develop a prediction model that can resemble the full load-settlement response of drilled shafts (bored piles) subjected to axial loading. The developed RNN model was calibrated and validated using several in situ full-scale pile load tests, as well as cone penetration test (CPT) data. The results indicate that the developed RNN model has the ability to reliably predict the load-settlement response of axially loaded drilled shafts and can thus be used by geotechnical engineers for routine design practice. [ABSTRACT FROM AUTHOR]

This paper proposes the use of online retrained metamodels for the optimization of water distribution system (WDS) design. In these metamodels, artificial neural networks (ANNs) are used to replace the full hydraulic and water quality simulation models and differential evolution (DE) is utilized to carry out the optimization. The ANNs in the proposed online DE-ANN model are retrained periodically during the optimization in order to improve their approximation to the appropriate portion of the search space. In addition, a local search strategy is used to further polish the final solution obtained by the online DE-ANN model. Three case studies are used to verify the effectiveness of the proposed online retrained DE-ANN model for which both hydraulic and water quality constraints are considered. In order to enable a performance comparison, a model in which a DE is combined with a full hydraulic and water quality simulation model (DE-EPANET2.0) and an offline DE-ANN model (ANNs are trained only once at the beginning of optimization) are established and applied to each case study. The results obtained show that the proposed online retrained DE-ANN model consistently outperforms the offline DE-ANN model for each case study in terms of efficiency and solution quality. Compared with the DE-EPANET2.0 model, the proposed online DE-ANN model exhibits a substantial improvement in computational efficiency, while still producing reasonably good quality solutions. [ABSTRACT FROM AUTHOR]

This paper presents the application of an improved particle swarm optimization (PSO) technique for training an artificial neural network (ANN) to predict water levels for the Heshui Watershed, China. Daily values of rainfall and water levels from 1988 to 2000 were first analyzed using ANNs trained with the conjugate gradient, gradient descent, and Levenberg-Marquardt neural network (LM-NN) algorithms. The best results were obtained from the LM-NN, and these results were then compared with those from PSO-based ANNs, including the conventional PSO neural network (CPSONN) and the improved PSO neural network (IPSONN) with passive congregation. The IPSONN algorithm improves PSO convergence by using the selfish herd concept in swarm behavior. The results show that the PSO-based ANNs performed better than the LM-NN. For models run using a single parameter (rainfall) as input, the root mean square error (RMSE) of the testing data set for IPSONN was the lowest (0.152 m) compared to those for CPSONN (0.161 m) and LM-NN (0.205 m). For multiparameter (rainfall and water level) inputs, the RMSE of the testing data set for IPSONN was also the lowest (0.089 m) compared to those for CPSONN (0.105 m) and LM-NN (0.145 m). The results also indicate that the LM-NN model performed poorly in predicting the low and peak water levels in comparison to the PSO-based ANNs. Moreover, the IPSONN model was superior to CPSONN in predicting extreme water levels. Lastly, IPSONN had a quicker convergence rate compared to CPSONN. [ABSTRACT FROM AUTHOR]

In this paper, the feasibility of developing and using artificial neural networks (ANNs) for modeling the monotonic behaviors of various angular and rounded rockfill materials is investigated. The database used for development of the ANN models is comprised of a series of 82 large-scale, drained triaxial tests. The deviator stress-volumetric strain versus axial strain behaviors were first simulated by using ANNs. A feedback model using multilayer perceptrons for predicting drained behavior of rockfill materials was developed in the MATLAB environment, and the optimal ANN architecture was obtained by a trial-and-error approach in accordance with error indexes and real data. Reasonable agreement between the simulated behaviors and the test results was observed, indicating that the ANNs are capable of capturing the behavior of rockfill materials. The ability of ANNs to predict the constitutive hardening-soil model parameters, residual deviator stresses, and corresponding volumetric strain was also investigated. Moreover, the generalization capability of ANNs was also used to check the effects of items not tested, such as dry density, grain-size distributions, and Los Angeles abrasion. [ABSTRACT FROM AUTHOR]

A fully automated data-driven methodology for the detection of pipe bursts and other events that induce similar abnormal pressure/flow variations (e.g., unauthorized consumptions) at the district metered area (DMA) level has been recently developed by the authors. This methodology works by simultaneously analyzing the data coming on-line from all the pressure and/or flow sensors deployed in a DMA. It makes synergistic use of several self-learning artificial intelligence (AI) and statistical techniques. These include (1) wavelets for the de-noising of the recorded pressure/flow signals; (2) artificial neural networks (ANNs) for the short-term forecasting of pressure/flow signal values; (3) statistical process control (SPC) techniques for the short-term and long-term analysis of the burst/other event-induced pressure/flow variations; and (4) Bayesian inference systems (BISs) for inferring the probability that a pipe burst/other event has occurred in the DMA being studied, raising the corresponding detection alarms, and provide information useful for performing event diagnosis. This paper focuses on the (re)calibration of the above detection methodology with the aim of improving the forecasting performances of the ANN models and the classification performances of the BIS used to raise the detection alarms (i.e., DMA-level BIS). This is achieved by using (1) an Evolutionary Algorithm optimization strategy for selecting the best ANN input structures and related parameter values to be used for training the ANN models, and (2) an Expectation Maximization strategy for (re)calibrating the values in the conditional probability tables (CPTs) of the DMA-level BIS. The (re)calibration procedure is tested on a case study involving several DMAs in the U.K. with real-life pipe bursts/other events, engineered pipe burst events (i.e., simulated by opening fire hydrants), and synthetic pipe burst events (i.e., simulated by arbitrarily adding 'burst flows' to an actual flow signal). The results obtained illustrate that the new (re)calibration procedure improves the performance of the event detection methodology in terms of increased detection speed and reliability. [ABSTRACT FROM AUTHOR]