Search

This paper develops a hierarchical Bayesian model (HBM) that integrates the physical knowledge and the test data to predict the small-strain shear modulus Gmax for a target sand type. The limited target-specific data are combined with the abundant generic data through a hierarchical structure so that the variability of Gmax within one sand type and across different sand types can be captured. The hyperparameters that characterize the same underlying distribution of physical model parameters across all the sand types are first estimated from the abundant generic data. The model parameters for the new sand type are then updated as the limited site-specific data become available. The approach is illustrated using a generic database and two real examples not covered by the generic database. Multiple possible hierarchical models are compared in terms of model complexity and goodness-of-fit. The results show that the hierarchical modeling of small-strain shear modulus data is reasonable and necessary. The hierarchical model can provide less biased and more accurate predictions of Gmax compared to the commonly used complete pooling model, especially in cases where the site-specific data are quite different from the overall average of the generic database. [ABSTRACT FROM AUTHOR]

Understanding the variation of geological interfaces plays a crucial role in the analysis and design of infrastructure systems. Generally, there are two classes of techniques for predicting geological interfaces, for example, interpolation/regression-based techniques and machine-learning-based techniques. In this paper, a Multi-scale Meta-learning Model (M3) methodology is proposed. The new methodology improves the current state-of-the-art techniques by fusing two levels of information: (i) generic characteristics of the sampling locations, for example, coordinates, and (ii) location-specific characteristics, for example, local-scale predictions. The implementation starts from using an array of classic interpolation/regression-based techniques as base learners to provide first-level predictions at a local scale. These predictions are then combined with generic characteristics to train a meta-learner following the stacking ensemble learning framework. In this manner, the location-specific information from the base learners can be simultaneously considered with the generic information in the training process. The variation of rockhead elevation is predicted using the M3 methodology and a comprehensive borehole dataset in Singapore. A detailed comparative study involving several existing methods is also carried out to rigorously validate the M3 methodology. The results show that the M3 methodology achieves 20% improvement in the model performance compared to existing methods, indicating its promising potential in geotechnical site characterization. [ABSTRACT FROM AUTHOR]

The ability of three methods to predict the load–settlement response of rock-socketed piles was examined in a recent paper. The piles considered were complete piles in which both the side and base of the piles contribute to the total resistance. For a range of reasons, it is sometimes necessary to consider rock-socketed piles for which the resistance is provided only by the side of the socket. This paper extends the earlier paper by comparing predictions for side-resistance-only piles with the load–settlement response of a suite of full-scale side-resistance-only pile tests. Only two of the methods of the earlier paper are used for this comparison, as the third method could not decouple the two components of resistance. It is demonstrated that one of the methods produces reasonable predictions of performance, whereas the other appears to underpredict performance. [ABSTRACT FROM AUTHOR]

This review paper presents the current state-of-the-art pertains to water pipe failure prediction and risk assessment, published in the last ten years (2009–2019). This paper has been motivated by the lack of comprehensive review articles that integrates water network failure and risk modeling. Some of the current practices reviewed the pipe condition and its failure. Others focused on the statistical prediction models, whereas the rest outlined failure prediction models of large diameter mains only. The mainstream of the current practice, highlighted in this paper characterizes the structural deterioration and failure rates using various statistical techniques, whereas the remainder of research covers a proliferation of machine learning and soft computing applications to forecast and model the pipeline risk of failure. The review offers descriptions of the models together with their proposed methodologies, algorithms and equations, contributions and drawbacks, comparisons and critiques, and types of data used to develop the models using the bibliographic review method. Finally, future work and research challenges are recommended to assist the civil engineering research community in setting a clear agenda for the upcoming research. [ABSTRACT FROM AUTHOR]

This paper describes a study focusing on performance evaluation of road weather information system (RWIS) surface temperature forecasts. To identify the factors influencing the accuracy of forecasts, five research hypotheses regarding climate patterns, geographical attributes, seasonal variations, time of day, and forecast horizon are proposed. Observation and forecast data were provided by four North American provinces, and were processed and stratified by station, hour, and month to quantify their effects. The findings based on two statistical measures; namely, mean absolute error and percent of acceptable forecasts suggested that the maritime climate group had the highest correspondence while the mixed climate group had the lowest. The geographical attributes showed that the forecasting performance of coastal maritime regions was found to have a negative correlation to the distance from nearby large bodies of water. Forecasts for the daytime were found to be less accurate than for the nighttime and deteriorate quickly as the forecasting horizon increased. [ABSTRACT FROM AUTHOR]

Vehicle–bridge collisions (VBCs) can compromise the safety of road users and cause major economic losses. This paper proposes and applies a methodology to investigate such events in Quebec. Relevant data have been collected from various sources and merged to provide a comprehensive database of VBCs that occurred in Quebec between 2000 and 2016. The developed database was used to carry out statistical analyses highlighting the main factors characterizing VBCs, such as vehicle's body type, bridge dimensions, prescribed speed limit, road configuration, road surface condition and lighting. The compiled database was georeferenced in an upgradable map that can be used efficiently to visualize the distribution and evolution of VBCs over a given region of Quebec. A VBC regression model was also developed based on k-fold cross-validation. The proposed model can be updated regularly as new VBCs are reported and then used to identify bridges most likely to be affected by VBCs or prioritize actions to reduce the potential consequences. [ABSTRACT FROM AUTHOR]

In many countries, airborne laser scanning (ALS) inventories are implemented to produce predictions for stand-level forest attributes. Nevertheless, mature stands are usually field-visited prior to clear-cutting, so some measurements can be conducted on these stands to calibrate the ALS-based predictions. In this paper, we developed a seemingly unrelated multivariate mixed-effects model system that includes component models for basal area, merchantable volume, and sawlog volume for 225 m2 cells. We used ALS data and accurately positioned cut-to-length harvester observations from clear-cut stands dominated by Norway spruce (Picea abies (L.) Karst.). Our aim was to study the effect of 1–10 local angle-gauge basal area measurements on the accuracy of predicted merchantable and sawlog volumes. A seemingly unrelated mixed-effect model system was fitted to estimate cross-model correlations in residuals and random effects, which were then utilized to predict all the random effects of the system for calibrated stand-level predictions. The 10 angle-gauge plots decreased the relative root mean square error (RMSE%) of the basal area and merchantable volume predictions from 16.8% to 10.5% and from 15.8% to 11.9%, respectively. Cross-model correlations of the stand effects of sawlog volume with the other responses were low; therefore, the initial RMSE% of ∼22% was decreased only marginally by the calibration. [ABSTRACT FROM AUTHOR]

We describe in this paper the quantum chromodynamics prediction to calculate the strong coupling constant by using event shape variables as well as momentum spectra. By fitting the dispersive model and employing our parameters on event shape distribution, we obtain the perturbative value of --> α s (M Z 0 ) --> = 0.1305 ± 0.0474 and also the non-perturbative value of α0 = 0.5246 ± 0.0516 GeV for electron–proton interactions. Next, by using momentum spectra for the same interactions, we obtain αs = 0.1572 ± 0.029. Our values in both methods are consistent with those obtained from electron–positron annihilations measured previously. When we find coupling constant for different flavours, we observe that they do not affect our results considerably. This is in accordance with quantum chromodynamics theory. All these features will be explained in the main text. [ABSTRACT FROM AUTHOR]

Accurate estimation of construction production, which is composed of productivity and unit costs, allows construction planners and managers to have excellent control over current processes and to correctly predict the production of similar projects in the future. Due to the need for accurate production estimation, selection of the appropriate construction technology is a critical factor in the success of a project. This paper presents a methodology for developing a model capable of predicting productivity and unit costs using several procedures, such as actual data collection, input data generation using construction simulation, and multiple regression analysis. An earthmoving operation was analyzed to estimate the proposed methodology’s prediction of construction production. A global positioning system (GPS)-based earthmoving system was selected as the new construction technology to be compared with the conventional system, to evaluate the decision-making process at a jobsite. The proposed methodology is expected to provide users with a basis for selecting appropriate technology. The case study presented in this paper demonstrates how to utilize the proposed methodology and analyze its predicted results. L'estimation précise de la production de construction, qui est composée de la productivité et des coûts unitaires, permet aux pilotes et gestionnaires de construction d'avoir un contrôle excellent sur les processus en cours et de prévoir correctement la production de projets similaires à l'avenir. En raison de la nécessité d'une estimation précise de production, le choix de la technologie de construction appropriée est un facteur critique dans la réussite d'un projet. Cet article présente une méthodologie pour développer un modèle capable de prédire la productivité et le coût unitaire en utilisant plusieurs procédures, telles que la collecte de données réelles, la génération de données d'entrée par simulation de construction et l'analyse de régression multiple. Une opération de terrassement a été analysée pour estimer la prédiction de la production de construction par la méthodologie proposée. Un engin de terrassement muni de système de positionnement global (« GPS ») a été choisi comme la nouvelle technologie de construction à comparer avec le système conventionnel pour évaluer le processus de prise de décision sur un chantier. La méthodologie proposée est censée fournir aux utilisateurs une base pour la sélection de la technologie appropriée. L'étude de cas présentée dans cet article montre comment utiliser la méthodologie proposée et analyser ses résultats prédits. [ABSTRACT FROM AUTHOR]

This paper examines the importance of the unit of analysis selected for trip generation modelling when the model estimation data are collected in a household travel survey. The paper reviews the literature on the arguments made for the use of the "individual" or the "household" as the unit of analysis in trip production modelling, and then through a statistical exposition it determines what should be the appropriate unit of analysis. An empirical test of the forecast performance of household- and person-trip generation models is conducted using data collected in a household-travel-behaviour survey in the Greater Toronto Area of Canada. The paper concludes that the household is theoretically the preferable analysis unit to use in trip production modelling when the model estimation data are collected in a household travel survey in which the household is the sampling unit. The empirical test indicates that household-trip generation models yield predictions of trips at the household and traffic zone level, respectively, that are marginally more accurate than those yielded by person-trip generation models. [ABSTRACT FROM AUTHOR]

Predictive water main break models can assist municipalities in prioritizing the replacement and rehabilitation of water mains. The aim of the paper is to develop an artificial neural network (ANN) model to forecast water main breaks in the water distribution network of the City of Kingston, Ontario, Canada. The ANN model includes variables of diameter, age, length, and soil type to forecast breaks. Historical break data from the 1998 to 2011 period is used to develop the ANN model and forecast pipe breaks over a 5 year planning period. The mean square error, receiver operating characteristics curves, and a confusion matrix are used to evaluate the ANN model training and testing. The trained neural network correctly classified 85% of the data set at the training, validation, and testing stages. Model forecasts showed lower pipe break rates in Kingston West, Kingston Central, and Kingston East. The reduction in break rate in the Kingston system was attributed to the removal of old pipes, and the favourable performance of pipes that are in the usage phase of their life cycle. The ANN model provided Utilities Kingston with a tool to assist them in the planning and management of their water main rehabilitation program. [ABSTRACT FROM AUTHOR]

In the last few decades, numerous methods have been developed for predicting the axial capacity of pile foundations. Among the available methods, the cone penetration test (CPT)-based models have been shown to give better predictions in many situations. This can be attributed to the fact that CPT-based methods have been developed in accordance with the CPT results, which have been found to yield more reliable soil properties; hence, more accurate axial pile capacity predictions. In this paper, one of the most commonly used artificial intelligence techniques, i.e., artificial neural networks (ANNs), is utilized in an attempt to develop artificial neural network (ANN) models that provide more accurate axial capacity predictions for driven piles and drilled shafts. The ANN models are developed using data collected from the literature and comprise 80 driven pile and 94 drilled-shaft load tests, as well as CPT results. The predictions from the ANN models are compared with those obtained from the most commonly used available CPT-based methods, and statistical analyses are carried out to rank and evaluate the performance of the ANN models and CPT methods. To facilitate the use of the developed ANN models, they are translated into simple design equations suitable for hand calculations. [ABSTRACT FROM AUTHOR]

Biological invasions continue to accelerate, and there is a need for closer integration between invasive species research and on-the-ground management. In many regions, aquatic invasive species have established isolated populations, but have not yet spread to many sites that provide suitable habitat. In the Laurentian Great Lakes region, several Great Lakes invaders such as zebra mussel (Dreissena polymorpha), rainbow smelt (Osmerus mordax), and spiny water flea (Bythotrephes longimanus) are currently undergoing secondary spread to the smaller inland lakes and streams. This paper describes recent advances in forecasting the secondary spread of aquatic invasive species and presents a framework for assessing vulnerability of inland waters based on explicit assessment of three distinct aspects of biological invasions: colonization, site suitability, and adverse impact. In many cases, only a fraction of lakes on the landscape are vulnerable to specific invasive species, highlighting the potential application of this type of research for improving invasive species management. Effective application to on-the-ground resource management will require that research aimed at assessing site vulnerability be translated into management tools. [ABSTRACT FROM AUTHOR]

This paper evaluates the accuracy of three commonly used models that predict accidents on two-lane, rural, arterial highways. The retrospective evaluation compared model outputs with empirical collision results for a sample of highway sections in the Province of New Brunswick. The analysis determined historical accident rates, identified key predictive variables, and compared the observed results with estimates from each safety model. All three models were found to significantly overestimate accident frequencies on the highway sections under study. The model generally employed in New Brunswick, MicroBENCOST, was found to yield the highest errors in estimated collisions. These findings suggest that the benefits from accident reduction are generally overestimated on highway improvement projects analyzed with these accident prediction models. Cet article évalue la précision de trois modèles couramment utilisés pour la prédiction d'accidents sur les autoroutes à deux voies; rurales et les avenues. L’évaluation rétrospective à comparé les résultats des modèles aux résultats empirique de collision pour un échantillon de sections d'autoroutes dans la province du Nouveau-Brunswick. L’analyse a permis de déterminer le taux historique d'accidents, a identifié les principales variables prédictives et à comparé les résultats observés aux estimations de chaque modèle de sécurité. Il a été trouvé que les trois modèles sur-estiment significativement les fréquences d’accidents sur les sections d’autoroutes étudiées. Il a été trouvé que le modèle généralement employé au Nouveau-Brunswick, MicroBENCOST, produit les erreurs les plus élevées pour l’estimation des collisions. Ces résultats suggèrent que les avantages dûs à la réduction des accidents sont généralement sur-estimés dans les projets d'amélioration des routes analysés avec ces modèles de prédiction des accidents. [ABSTRACT FROM AUTHOR]

Artificial neural network (ANN) technology has evolved from the experimental stage into actual industrial applications. To achieve this significant transition, careful planning and adjustment are required. This paper illustrates such an example in the water treatment industry. The project objective is to upgrade the ANN models from a previous research project and install the system on-line in the Rossdale Water Treatment Plant in Edmonton, Alberta, Canada, to forecast raw water colour one day ahead. The article discusses the important issues and techniques to upgrade the neural network model to the actual application. Furthermore, sufficient communication is also required between the designers and the users to address the applicability and user friendly issues in model implementation. Failure in communication can render the whole process ineffective. Possible improvements are also recommended for the future on-line applications. [ABSTRACT FROM AUTHOR]