Search results for: accidents prediction models (APMs)

Authors: Samuel Opeyemi Williams, Razali Adul Hamid, Mohd Saidin Misman, Dominic Ileyemi Ajayi, Taki Eddine Seghier

Abstract:

Construction industry is famous with hazardous and high-risk environment with operatives facing a greater risk of work-related fatality or injury than operatives in other industries. It is characterised with different types of accident, ranging from electrocution, trip and slip, fall from height, struck-by, explosion, trench collapse, to scaffolding accidents, with each type being caused by different factors. However, accidents are unplanned, unforeseeable and unexpected events, but regardless of the high-risk nature of the industry, accidents are preventable. The aim of the paper is to determine the extent of the utilization of the safety measures, as well as identifying the factors underlining the non-usage. A research methodology consisting of a literature review was used to identify the types and causes of site accidents, while a well-structured questionnaire was used to determine the level of the usage of safety measures on site. The data were analysed with the results revealing the use of protective clothing, safety helmet, first aid, protective shoe, safety belt, and face shield to aid safety of workers, as well as ascribing non-usage of safety measures to cost, ignorance, lack of experts and non-inclusion in contract document. Recommendations are included in the paper suggesting the enforcement of the utilization of safety measures in reducing the spate of accident occurrence on construction sites.

Keywords: construction industry, safety measures, accident, prevention

Procedia PDF Downloads 293

Authors: Matevž Breška, Iztok Peruš, Vlado Stankovski

Abstract:

Systematic overview of existing Ground Motion Prediction Equations (GMPEs) has been published by Douglas. The number of earthquake recordings that have been used for fitting these equations has increased in the past decades. The current PF-L database contains 3550 recordings. Since the GMPEs frequently model the peak ground acceleration (PGA) the goal of the present study was to refit a selection of 44 of the existing equation models for PGA in light of the latest data. The algorithm Levenberg-Marquardt was used for fitting the coefficients of the equations and the results are evaluated both quantitatively by presenting the root mean squared error (RMSE) and qualitatively by drawing graphs of the five best fitted equations. The RMSE was found to be as low as 0.08 for the best equation models. The newly estimated coefficients vary from the values published in the original works.

Keywords: Ground Motion Prediction Equations, Levenberg-Marquardt algorithm, refitting PF-L database, peak ground acceleration

Procedia PDF Downloads 430

Authors: Amartya Hatua, Trung Nguyen, Andrew Sung

Abstract:

In this paper, we study the information diffusion process on Twitter as a multivariate time series problem. Our model concerns three measures (volume, network influence, and sentiment of tweets) based on 10 features, and we collected 27 million tweets to build our information diffusion time series dataset for analysis. Then, different time series clustering techniques with Dynamic Time Warping (DTW) distance were used to identify different patterns of information diffusion. Finally, we built the information diffusion prediction models for new hashtags which comprise two phrases: The first phrase is recognizing the pattern using k-NN with DTW distance; the second phrase is building the forecasting model using the traditional Autoregressive Integrated Moving Average (ARIMA) model and the non-linear recurrent neural network of Long Short-Term Memory (LSTM). Preliminary results of performance evaluation between different forecasting models show that LSTM with clustering information notably outperforms other models. Therefore, our approach can be applied in real-world applications to analyze and predict the information diffusion characteristics of selected topics or memes (hashtags) in Twitter.

Keywords: ARIMA, DTW, information diffusion, LSTM, RNN, time series clustering, time series forecasting, Twitter

Procedia PDF Downloads 364

Authors: Danilo López, Edwin Rivas, Fernando Pedraza

Abstract:

Currently, one of the major challenges in wireless networks is the optimal use of radio spectrum, which is managed inefficiently. One of the solutions to existing problem converges in the use of Cognitive Radio (CR), as an essential parameter so that the use of the available licensed spectrum is possible (by secondary users), well above the usage values that are currently detected; thus allowing the opportunistic use of the channel in the absence of primary users (PU). This article presents the results found when estimating or predicting the future use of a spectral transmission band (from the perspective of the PU) for a chaotic type channel arrival behavior. The time series prediction method (which the PU represents) used is ANFIS (Adaptive Neuro Fuzzy Inference System). The results obtained were compared to those delivered by the RNA (Artificial Neural Network) algorithm. The results show better performance in the characterization (modeling and prediction) with the ANFIS methodology.

Keywords: ANFIS, cognitive radio, prediction primary user, RNA

Procedia PDF Downloads 395

Authors: Ajai Singh

Abstract:

Simulation of hydrological processes at the watershed outlet through modelling approach is essential for proper planning and implementation of appropriate soil conservation measures in Damodar Barakar catchment, Hazaribagh, India where soil erosion is a dominant problem. This study quantifies the parametric uncertainty involved in simulation of stream flow using Soil and Water Assessment Tool (SWAT), a watershed scale model and Radial Basis Neural Network (RBNN), an artificial neural network model. Both the models were calibrated and validated based on measured stream flow and quantification of the uncertainty in SWAT model output was assessed using ‘‘Sequential Uncertainty Fitting Algorithm’’ (SUFI-2). Though both the model predicted satisfactorily, but RBNN model performed better than SWAT with R2 and NSE values of 0.92 and 0.92 during training, and 0.71 and 0.70 during validation period, respectively. Comparison of the results of the two models also indicates a wider prediction interval for the results of the SWAT model. The values of P-factor related to each model shows that the percentage of observed stream flow values bracketed by the 95PPU in the RBNN model as 91% is higher than the P-factor in SWAT as 87%. In other words the RBNN model estimates the stream flow values more accurately and with less uncertainty. It could be stated that RBNN model based on simple input could be used for estimation of monthly stream flow, missing data, and testing the accuracy and performance of other models.

Keywords: SWAT, RBNN, SUFI 2, bootstrap technique, stream flow, simulation

Procedia PDF Downloads 331

Authors: Ramesh Kumar Behera, Md. Izhar Hassan

Abstract:

Accident investigation and analysis are useful to identify deficiencies in plant, process, and management practices and formulate preventive strategies for injury elimination. In India and other parts of the world, industrial accidents are investigated to know the causes and also to fulfill legal compliances. However, findings of investigation are seldom used appropriately to strengthen Occupational Safety and Health (OSH) in expected lines. The mineral rich state of Odisha in eastern coast of India; known as a hub for Iron and Steel industries, witnessed frequent accidents during 2005-2009. This article based on study of 982 fatal ‘factory-accidents’ occurred in Odisha during the period 2001-2016, discusses the ‘turnaround-story’ resulting in reduction of fatal accident from 122 in 2009 to 45 in 2016. This paper examines various factors causing incidents; accident pattern in steel and chemical sector; role of climate and harsh weather conditions on accident causation. Software such as R, SQL, MS-Excel and Tableau were used for analysis of data. It is found that maximum fatality is caused due to ‘fall from height’ (24%); steel industries are relatively more accident prone; harsh weather conditions of summer increase chances of accident by 20%. Further, the study suggests that enforcement of partial work-restriction around lunch time during peak summer, screening and training of employees reduce accidents due to fall from height. The study indicates that learning from accident investigation and analysis can be used as a method to reduce work related accidents in the journey towards ‘Vision Zero’.

Keywords: accident investigation and analysis, fatal accidents in India, fall from height, vision zero

Procedia PDF Downloads 130

Authors: Emma Roberts

Abstract:

The Rome II Regulation has established a uniform regime of conflict of law rules across the European Union (except for Denmark) which determines the law applicable in non-contractual obligations disputes. It marks a significant development towards the Europeanization of private international law and aims to provide the most appropriate connecting factors to achieve both legal certainty and justice in individual cases. Many non-contractual obligations are recognised to present such distinct factors that, to achieve these aims, a special rule is provided for determining the applicable law in cases in respect of product liability and environmental torts, for example. Throughout the legislative process, the European Parliament sought to establish a separate rule for road traffic accidents, recognising that these cases too present such novel situations that a blanket application of a lex loci damni approach would not provide an appropriate answer. Such attempts were rejected and, as a result, cases arising out of road traffic accidents are subject to the Regulation’s general lex loci damni rule along with its escape clause and limited exception. This paper offers a critique of the Regulation’s response to cross-border road traffic accident cases. In England and Wales, there have been few cases that have applied the Regulation’s provisions to date, but significantly the majority of such cases are in respect of road traffic accidents. This paper examines the decisions in those cases and challenges the legislators’ decision not to provide a special rule for such incidences. Owing to the diversity in compensation systems globally, applying the Regulation’s general rule to cases of road traffic accidents – given the breadth of matters that are to be subject to the lex cause – cannot ensure an outcome that provides ‘justice in individual cases’ as is assured by the Regulation's recitals. Not only does this paper suggest that the absence of a special rule for road traffic accidents means that the Regulation fails to achieve one of its principal aims, but it further makes out a compelling case for the legislative body of the European Union to implement a corrective instrument.

Keywords: accidents abroad, applicable law, cross-border torts, non-contractual obligations, road traffic accidents

Procedia PDF Downloads 235

Authors: E. A. Mlybari, M. S. Elbisy, A. H. Alshahri, O. M. Albarakati

Abstract:

Sea level rise threatens to increase the impact of future storms and hurricanes on coastal communities. Accurate sea level change prediction and supplement is an important task in determining constructions and human activities in coastal and oceanic areas. In this study, support vector machines (SVM) is proposed to predict daily tidal levels along the Jeddah Coast, Saudi Arabia. The optimal parameter values of kernel function are determined using a genetic algorithm. The SVM results are compared with the field data and with back propagation (BP). Among the models, the SVM is superior to BPNN and has better generalization performance.

Keywords: tides, prediction, support vector machines, genetic algorithm, back-propagation neural network, risk, hazards

Procedia PDF Downloads 440

Authors: Georges Ghazi, Magali Gelhaye, Ruxandra Botez

Abstract:

Over the years, the Flight Management System (FMS) has experienced a continuous improvement of its many features, to the point of becoming the pilot’s primary interface for flight planning operation on the airplane. With the assistance of the FMS, the concept of distance and time has been completely revolutionized, providing the crew members with the determination of the optimized route (or flight plan) from the departure airport to the arrival airport. To accomplish this function, the FMS needs an accurate Aircraft Performance Model (APM) of the aircraft. In general, APMs that equipped most modern FMSs are established before the entry into service of an individual aircraft, and results from the combination of a set of ordinary differential equations and a set of performance databases. Unfortunately, an aircraft in service is constantly exposed to dynamic loads that degrade its flight characteristics. These degradations endow two main origins: airframe deterioration (control surfaces rigging, seals missing or damaged, etc.) and engine performance degradation (fuel consumption increase for a given thrust). Thus, after several years of service, the performance databases and the APM associated to a specific aircraft are no longer representative enough of the actual aircraft performance. It is important to monitor the trend of the performance deterioration and correct the uncertainties of the aircraft model in order to improve the accuracy the flight management system predictions. The basis of this research lies in the new ability to continuously update an Aircraft Performance Model (APM) during flight using an adaptive lookup table technique. This methodology was developed and applied to the well-known Cessna Citation X business aircraft. For the purpose of this study, a level D Research Aircraft Flight Simulator (RAFS) was used as a test aircraft. According to Federal Aviation Administration the level D is the highest certification level for the flight dynamics modeling. Basically, using data available in the Flight Crew Operating Manual (FCOM), a first APM describing the variation of the engine fan speed and aircraft fuel flow w.r.t flight conditions was derived. This model was next improved using the proposed methodology. To do that, several cruise flights were performed using the RAFS. An algorithm was developed to frequently sample the aircraft sensors measurements during the flight and compare the model prediction with the actual measurements. Based on these comparisons, a correction was performed on the actual APM in order to minimize the error between the predicted data and the measured data. In this way, as the aircraft flies, the APM will be continuously enhanced, making the FMS more and more precise and the prediction of trajectories more realistic and more reliable. The results obtained are very encouraging. Indeed, using the tables initialized with the FCOM data, only a few iterations were needed to reduce the fuel flow prediction error from an average relative error of 12% to 0.3%. Similarly, the FCOM prediction regarding the engine fan speed was reduced from a maximum error deviation of 5.0% to 0.2% after only ten flights.

Keywords: aircraft performance, cruise, trajectory optimization, adaptive lookup tables, Cessna Citation X

Procedia PDF Downloads 235

Authors: Armana Huq, Wahidur Rahman, Sanwar Kader

Abstract:

Inland waterways in Bangladesh play an important role in providing comfortable and low-cost transportation. However, a maritime accident takes away many lives and creates unwanted hazards every year. This article deals with a comprehensive review of inland waterway accidents in Bangladesh. Additionally, it includes a comparative study between international and local inland research studies based on maritime accidents. Articles from inland waterway areas are analyzed in-depth to make a comprehensive overview of the nature of the academic work, accident and risk management process and different statistical analyses. It is found that empirical analysis based on the available statistical data dominates the research domain. For this study, major maritime accident-related works in the last four decades in Bangladesh (1981-2020) are being analyzed for preparing a bibliometric analysis. A study of maritime accidents of passenger's vessels during (1995-2005) indicates that the predominant causes of accidents in the inland waterways of Bangladesh are collision and adverse weather (77%), out of which collision due to human error alone stands (56%) of all accidents. Another study refers that the major causes of waterway accidents are the collision (60.3%) during 2005-2015. About 92% of this collision occurs due to direct contact with another vessel during this period. Rest 8% of the collision occurs by contact with permanent obstruction on waterway roots. The overall analysis of another study from the last 25 years (1995-2019) shows that one of the main types of accidents is collisions, with about 50.3% of accidents being caused by collisions. The other accident types are cyclone or storm (17%), overload (11.3%), physical failure (10.3%), excessive waves (5.1%), and others (6%). Very few notable works are available in testing or comparing the methods, proposing new methods for risk management, modeling, uncertainty treatment. The purpose of this paper is to provide an overview of the evolution of marine accident-related research domain regarding inland waterway of Bangladesh and attempts to introduce new ideas and methods to abridge the gap between international and national inland maritime-related work domain which can be a catalyst for a safer and sustainable water transportation system in Bangladesh. Another fundamental objective of this paper is to navigate various national maritime authorities and international organizations to implement risk management processes for shipping accident prevention in waterway areas.

Keywords: inland waterways, safety, bibliometric analysis, risk management, accidents

Procedia PDF Downloads 163

Authors: Nebila Lichiheb, LaToya Myles, William Pendergrass, Bruce Hicks, Dawson Cagle

Abstract:

Atmospheric transport of hazardous materials in urban areas is increasingly under investigation due to the potential impact on human health and the environment. In response to health and safety concerns, several dispersion models have been developed to analyze and predict the dispersion of hazardous contaminants. The models of interest usually rely on meteorological information obtained from the meteorological models of NOAA’s National Weather Service (NWS). However, due to the complexity of the urban environment, NWS forecasts provide an inadequate basis for dispersion computation in urban areas. A dense meteorological network in Washington, DC, called DCNet, has been operated by NOAA since 2003 to support the development of urban monitoring methodologies and provide the driving meteorological observations for atmospheric transport and dispersion models. This study focuses on the comparison of wind observations from the DCNet station on the U.S. Department of Commerce Herbert C. Hoover Building against the North American Mesoscale (NAM) model outputs for the period 2017-2019. The goal is to develop a simple methodology for modifying NAM outputs so that the dispersion requirements of the city and its urban area can be satisfied. This methodology will allow us to quantify the prediction errors of the NAM model and propose adjustments of key variables controlling dispersion model calculation.

Keywords: meteorological data, Washington D.C., DCNet data, NAM model

Procedia PDF Downloads 209

Authors: Masahiro Miyaji

Abstract:

When a driver happens to be involved in some traffic congestion or after traffic incidents, the driver may fall in a state of anger. State of anger may encounter decisive risk resulting in severer traffic accidents. Preventive safety function using driver’s psychosomatic state with regard to anger may be one of solutions which would avoid that kind of risks. Identifying driver’s anger state is important to create countermeasures to prevent the risk of traffic accidents. As a first step, this research figured out root cause of traffic incidents by means of using Internet survey. From statistical analysis of the survey, dominant psychosomatic states immediately before traffic incidents were haste, distraction, drowsiness and anger. Then, we replicated anger state of a driver while driving, and then, replicated it by means of using driving simulator on bench test basis. Six types of facial expressions including anger were introduced as alternative characteristics. Kohonen neural network was adopted to classify anger state. Then, we created a methodology to detect anger state of a driver in high accuracy. We presented a driving support safety function. The function adapts driver’s anger state in cooperation with an autonomous driving unit to reduce the number of traffic accidents. Consequently, e evaluated reduction rate of driver’s anger in the traffic accident. To validate the estimation results, we referred the reduction rate of Advanced Safety Vehicle (ASV) as well as Intelligent Transportation Systems (ITS).

Keywords: Kohonen neural network, driver’s anger state, reduction of traffic accidents, driver’s state adaptive driving support safety

Procedia PDF Downloads 327

Authors: Hui Zhang, Sherif Beskhyroun

Abstract:

Structural Health monitoring systems (SHM) are commonly used to identify and assess structural damage. In terms of damage detection, SHM needs to periodically collect data from sensors placed in the structure as damage-sensitive features. This includes abnormal changes caused by the strain field and abnormal symptoms of the structure, such as damage and deterioration. Currently, deploying sensors on a large scale in a building structure is a challenge. In this study, a highly stretchable strain sensors are used in this study to collect data sets of strain generated on the surface of full-size reinforced concrete (RC) frames under extreme cyclic load application. This sensing sheet can be switched freely between the test bending strain and the axial strain to achieve two different configurations. On this basis, the deep neural network prediction model of the frame beam and frame column is established. The training results show that the method can accurately predict the strain value and has good generalization ability. The two deep neural network prediction models will also be deployed in the SHM system in the future as part of the intelligent strain sensor system.

Keywords: strain sensing sheets, deep neural networks, strain measurement, SHM system, RC frames

Procedia PDF Downloads 66

Authors: Ali Ahmadi, Maryam Abbasi, Mohammad Mehdi Parsaei

Abstract:

Background: Stress in the workplace today is one of the most important public health concerns and a serious threat to the health of the workforce worldwide. Occupational stress can cause occupational events and reduce quality of life. As a result, it has a very undesirable impact on the performance of organizations, companies, and their human resources. This study aimed to evaluate the relationship between job stress dimensions and occupational accidents in industrial factories in Southwest Iran. Materials and Methods: This cross-sectional study was conducted among 200 workers in the summer of 2023 in the Southwest of Iran. To select participants, we used a convenience sampling method. The research tools in this study were the Health and Safety Executive (HSE) stress questionnaire with 35 questions and 7 dimensions and demographic information. A high score on this questionnaire indicates that there is low job stress and pressure. All workers completed the informed consent form. Univariate analysis was performed using chi-square and T-test. Multiple regression analysis was used to estimate the odds ratios (OR) and 95% confidence interval (CI) for the association of stress-related factors with job accidents in participants. Stata 14.0 software was used for analysis. Results: The mean age of the participants was 39.81(6.36) years. The prevalence of job accidents was 28.0% (95%CI: 21.0, 34.0). Based on the results of the multiple logistic regression with the adjustment of the effect of the confounding variables, one increase in the score of the demand dimension had a protective impact on the risk of job accidents(aOR=0.91,95%CI:0.85-0.95). Additionally, an increase in one of the scores of the managerial support (aOR=0.89, 95% CI: 0.83-0.95) and peer support (aOR=0.76, 95%CI: 0.67-87) dimensions was associated with a lower number of job accidents. Among dimensions, an increase in the score of relationship (aOR=0.89, 95%CI: 0.80-0.98) and change (aOR=0.86, 95%CI: 0.74-0.96) reduced the odds of the accident's occurrence among the workers by 11% and 16%, respectively. However, there was no significant association between role and control dimensions and the job accident (p>0.05). Conclusions: The results show that the prevalence of job accidents was alarmingly high. Our results suggested that an increase in scores of dimensions HSE questioners is significantly associated with a decrease the accident occurrence in the workplace. Therefore, planning to address stressful factors in the workplace seems necessary to prevent occupational accidents.

Keywords: HSE, Iran, job stress occupational accident, safety, occupational health

Procedia PDF Downloads 36

Authors: Yilin Liao, Hewen Li, Paula McConvey

Abstract:

Speed skaters often face a risk of concussion when they fall on the ice floor and impact crash mats during practices and competitive races. Several variables, including those related to the skater, the crash mat, and the impact position (body side/head/feet impact), are believed to influence the severity of the skater's concussion. While computer simulation modeling can be employed to analyze these accidents, the simulation process is time-consuming and does not provide rapid information for coaches and teams to assess the skater's injury risk in competitive events. This research paper promotes the exploration of the feasibility of using AI techniques for evaluating skater’s potential concussion severity, and to develop a fast concussion prediction tool using artificial neural networks to reduce the risk of treatment delays for injured skaters. The primary data is collected through virtual tests and physical experiments designed to simulate skater-mat impact. It is then analyzed to identify patterns and correlations; finally, it is used to train and fine-tune the artificial neural networks for accurate prediction. The development of the prediction tool by employing machine learning strategies contributes to the application of AI methods in sports science and has theoretical involvements for using AI techniques in predicting and preventing sports-related injuries.

Keywords: artificial neural networks, concussion, machine learning, impact, speed skater

Procedia PDF Downloads 69

Authors: Hy Dang, Bo Mei

Abstract:

The development of machine learning methods and techniques has opened doors for investigation in many areas such as medicines, economics, finance, etc. One active research area involving machine learning is stock market prediction. This research paper tries to consider multiple techniques and methods for stock movement prediction using historical price or price factors. The paper explores the effectiveness of some deep learning frameworks for forecasting stock. Moreover, an architecture (TimeStock) is proposed which takes the representation of time into account apart from the price information itself. Our model achieves a promising result that shows a potential approach for the stock movement prediction problem.

Keywords: classification, machine learning, time representation, stock prediction

Procedia PDF Downloads 117

Authors: Avi Harel

Abstract:

Most accidents are commonly attributed in hindsight to human errors, yet most methodologies for safety focus on technical issues. According to the Black Swan theory, this paradox is due to insufficient data about the ways systems fail. The article presents a study of the sources of errors, and proposes a methodology for utility-oriented design, comprising methods for coping with each of the sources identified. Accident analysis indicates that errors typically result from difficulties of operating in exceptional conditions. Therefore, following STAMP, the focus should be on preventing exceptions. Exception analysis indicates that typically they involve an improper account of the operational scenario, due to deficiencies in the system integration. The methodology proposes a model, which is a formal definition of the system operation, as well as principles and guidelines for safety-oriented system integration. The article calls to develop and integrate tools for recording and analysis of the system activity during the operation, required to implement validate the model.

Keywords: accidents, complexity, errors, exceptions, interaction, modeling, resilience, risks

Procedia PDF Downloads 176

Authors: A. Juhász, M. Fogarasi

Abstract:

Those causing fatal traffic accidents are traumatized, which negatively influences their cognitive functions and their personality. In order to clarify how much the trauma of causing a fatal accident effects their driving skills and personality traits, the results of a psychological aptitude and a personality test of drivers carelessly causing fatal accidents and of drivers not causing any accidents were compared separately. The sample (N = 354) consists of randomly selected drivers from the Transportation Aptitude and Examination Centre database who caused fatal accidents (Fatal group, n = 177) or did not cause accidents (Control group, n = 177). The aptitude tests were taken between 2014 and 2019. The comparison of the 2 groups was done according to 3 aspects: 1. Categories of aptitude (suitable, restricted, unsuited); 2. Categories of causes (ability, personality, ability and personality) within the restricted or unsuited (altogether: non-suitable subgroups); 3. Categories of ability and personality within the non-suitable subgroups regardless of the cause-category. Within ability deficiency, the two groups include those, whose ability factor is impaired or limited. This is also true in case of personality failure. Compared to the control group, the number of restricted drivers causing fatal accidents is significantly higher (p < .000) and the number of unsuited drivers is higher on a tendency-level (p = .06). Compared to the control group in the fatal non-suitable subgroup, the ratio of restricted suitability and the unsuitability due to ability factors is exclusively significantly lower (p < .000). The restricted suitability and the unsuitability due to personality factors are more significant in the fatal non-suitable subgroup (p < .000). Incapacity due to combination of ability and personality is also significantly higher in the fatal group (p = .002). Compared to the control group both ability and personality factors are also significantly higher in the fatal non-suitable subgroup (p < .000). Overall, the control group is more eligible for driving than drivers who have caused fatalities. The ability and personality factors are significantly higher in the case of fatal accident causers who are non-suitable for driving. Moreover the concomitance of ability and personality factors occur almost exclusively to drivers who caused fatal accidents. Further investigation is needed to understand the causes and how the aptitude test results for the fatal group could improve over time.

Keywords: aptitude, unsuited, fatal accident, ability, personality

Procedia PDF Downloads 120

Authors: Jairson Barbosa Rodrigues, Paulo Romero Martins Maciel, Germano Crispim Vasconcelos

Abstract:

This paper presents an investigation of the performance impacts regarding the variation of five factors (input data size, node number, cores, memory, and disks) when applying a distributed implementation of Naïve Bayes for text classification of a large Corpus on the Spark big data processing framework. Problem: The algorithm's performance depends on multiple factors, and knowing before-hand the effects of each factor becomes especially critical as hardware is priced by time slice in cloud environments. Objectives: To explain the functional relationship between factors and performance and to develop linear predictor models for time and cost. Methods: the solid statistical principles of Design of Experiments (DoE), particularly the randomized two-level fractional factorial design with replications. This research involved 48 real clusters with different hardware arrangements. The metrics were analyzed using linear models for screening, ranking, and measurement of each factor's impact. Results: Our findings include prediction models and show some non-intuitive results about the small influence of cores and the neutrality of memory and disks on total execution time, and the non-significant impact of data input scale on costs, although notably impacts the execution time.

Keywords: big data, design of experiments, distributed machine learning, natural language processing, spark

Procedia PDF Downloads 86

Authors: Siva K. Bathina, Sudheer Siddapureddy

Abstract:

Pool fires are formed when the flammable liquid accidentally spills on the ground or water and ignites. Pool fire is a kind of buoyancy-driven and diffusion flame. There have been many pool fire accidents caused during processing, handling and storing of liquid fuels in chemical and oil industries. Such kind of accidents causes enormous damage to property as well as the loss of lives. Pool fires are complex in nature due to the strong interaction among the combustion, heat and mass transfers and pyrolysis at the fuel surface. Moreover, the experimental study of such large complex fires involves fire safety issues and difficulties in performing experiments. In the present work, large eddy simulations are performed to study such complex fire scenarios using fire dynamic simulator. A 1 m diesel pool fire is considered for the studied cases, and diesel is chosen as it is most commonly involved fuel in fire accidents. Fire simulations are performed by specifying two different boundary conditions: one the fuel is in liquid state and pyrolysis model is invoked, and the other by assuming the fuel is initially in a vapor state and thereby prescribing the mass loss rate. A domain of size 11.2 m × 11.2 m × 7.28 m with uniform structured grid is chosen for the numerical simulations. Grid sensitivity analysis is performed, and a non-dimensional grid size of 12 corresponding to 8 cm grid size is considered. Flame properties like mass burning rate, irradiance, and time-averaged axial flame temperature profile are predicted. The predicted steady-state mass burning rate is 40 g/s and is within the uncertainty limits of the previously reported experimental data (39.4 g/s). Though the profile of the irradiance at a distance from the fire along the height is somewhat in line with the experimental data and the location of the maximum value of irradiance is shifted to a higher location. This may be due to the lack of sophisticated models for the species transportation along with combustion and radiation in the continuous zone. Furthermore, the axial temperatures are not predicted well (for any of the boundary conditions) in any of the zones. The present study shows that the existing models are not sufficient enough for modeling blended fuels like diesel. The predictions are strongly dependent on the experimental values of the soot yield. Future experiments are necessary for generalizing the soot yield for different fires.

Keywords: burning rate, fire accidents, fire dynamic simulator, pyrolysis

Procedia PDF Downloads 171

Authors: Amin Salighehdar, Ziwen Ye, Mingzhe Liu, Ionut Florescu, Alan F. Blumberg

Abstract:

Storm surge is an abnormal water level caused by a storm. Accurate prediction of a storm surge is a challenging problem. Researchers developed various ensemble modeling techniques to combine several individual forecasts to produce an overall presumably better forecast. There exist some simple ensemble modeling techniques in literature. For instance, Model Output Statistics (MOS), and running mean-bias removal are widely used techniques in storm surge prediction domain. However, these methods have some drawbacks. For instance, MOS is based on multiple linear regression and it needs a long period of training data. To overcome the shortcomings of these simple methods, researchers propose some advanced methods. For instance, ENSURF (Ensemble SURge Forecast) is a multi-model application for sea level forecast. This application creates a better forecast of sea level using a combination of several instances of the Bayesian Model Averaging (BMA). An ensemble dressing method is based on identifying best member forecast and using it for prediction. Our contribution in this paper can be summarized as follows. First, we investigate whether the ensemble models perform better than any single forecast. Therefore, we need to identify the single best forecast. We present a methodology based on a simple Bayesian selection method to select the best single forecast. Second, we present several new and simple ways to construct ensemble models. We use correlation and standard deviation as weights in combining different forecast models. Third, we use these ensembles and compare with several existing models in literature to forecast storm surge level. We then investigate whether developing a complex ensemble model is indeed needed. To achieve this goal, we use a simple average (one of the simplest and widely used ensemble model) as benchmark. Predicting the peak level of Surge during a storm as well as the precise time at which this peak level takes place is crucial, thus we develop a statistical platform to compare the performance of various ensemble methods. This statistical analysis is based on root mean square error of the ensemble forecast during the testing period and on the magnitude and timing of the forecasted peak surge compared to the actual time and peak. In this work, we analyze four hurricanes: hurricanes Irene and Lee in 2011, hurricane Sandy in 2012, and hurricane Joaquin in 2015. Since hurricane Irene developed at the end of August 2011 and hurricane Lee started just after Irene at the beginning of September 2011, in this study we consider them as a single contiguous hurricane event. The data set used for this study is generated by the New York Harbor Observing and Prediction System (NYHOPS). We find that even the simplest possible way of creating an ensemble produces results superior to any single forecast. We also show that the ensemble models we propose generally have better performance compared to the simple average ensemble technique.

Keywords: Bayesian learning, ensemble model, statistical analysis, storm surge prediction

Procedia PDF Downloads 289

Authors: Jacob J. Shila, Xiaoyu O. Wu

Abstract:

As the applications of unmanned aerial vehicles (UAV) continue to increase across the world, it is critical to understand the factors that contribute to incidents and accidents associated with these systems. Given the variety of daily applications that could utilize the operations of the UAV (e.g., medical, security operations, construction activities, landscape activities), the main discussion has been how to safely incorporate the UAV into the national airspace system. The types of UAV incidents being reported range from near sightings by other pilots to actual collisions with aircraft or UAV. These incidents have the potential to impact the rest of aviation operations in a variety of ways, including human lives, liability costs, and delay costs. One of the largest causes of these incidents cited is the human factor; other causes cited include maintenance, aircraft, and others. This work investigates the key human factors associated with UAV incidents. To that end, the data related to UAV incidents that have occurred in the United States is both reviewed and analyzed to identify key human factors related to UAV incidents. The data utilized in this work is gathered from the Federal Aviation Administration (FAA) drone database. This study adopts the human factor analysis and classification system (HFACS) to identify key human factors that have contributed to some of the UAV failures to date. The uniqueness of this work is the incorporation of UAV incident data from a variety of applications and not just military data. In addition, identifying the specific human factors is crucial towards developing safety operational models and human factor guidelines for the UAV. The findings of these common human factors are also compared to similar studies in other countries to determine whether these factors are common internationally.

Keywords: human factors, incidents and accidents, safety, UAS, UAV

Procedia PDF Downloads 217

Authors: Xiangtuo Chen, Paul-Henry Cournéde

Abstract:

Crop yield prediction is a paramount issue in agriculture. The main idea of this paper is to find out efficient way to predict the yield of corn based meteorological records. The prediction models used in this paper can be classified into model-driven approaches and data-driven approaches, according to the different modeling methodologies. The model-driven approaches are based on crop mechanistic modeling. They describe crop growth in interaction with their environment as dynamical systems. But the calibration process of the dynamic system comes up with much difficulty, because it turns out to be a multidimensional non-convex optimization problem. An original contribution of this paper is to propose a statistical methodology, Multi-Scenarios Parameters Estimation (MSPE), for the parametrization of potentially complex mechanistic models from a new type of datasets (climatic data, final yield in many situations). It is tested with CORNFLO, a crop model for maize growth. On the other hand, the data-driven approach for yield prediction is free of the complex biophysical process. But it has some strict requirements about the dataset. A second contribution of the paper is the comparison of these model-driven methods with classical data-driven methods. For this purpose, we consider two classes of regression methods, methods derived from linear regression (Ridge and Lasso Regression, Principal Components Regression or Partial Least Squares Regression) and machine learning methods (Random Forest, k-Nearest Neighbor, Artificial Neural Network and SVM regression). The dataset consists of 720 records of corn yield at county scale provided by the United States Department of Agriculture (USDA) and the associated climatic data. A 5-folds cross-validation process and two accuracy metrics: root mean square error of prediction(RMSEP), mean absolute error of prediction(MAEP) were used to evaluate the crop prediction capacity. The results show that among the data-driven approaches, Random Forest is the most robust and generally achieves the best prediction error (MAEP 4.27%). It also outperforms our model-driven approach (MAEP 6.11%). However, the method to calibrate the mechanistic model from dataset easy to access offers several side-perspectives. The mechanistic model can potentially help to underline the stresses suffered by the crop or to identify the biological parameters of interest for breeding purposes. For this reason, an interesting perspective is to combine these two types of approaches.

Keywords: crop yield prediction, crop model, sensitivity analysis, paramater estimation, particle swarm optimization, random forest

Procedia PDF Downloads 209

Authors: Sophia Liang Zhou

Abstract:

For investors, businesses, consumers, and governments, an accurate assessment of future housing prices is crucial to critical decisions in resource allocation, policy formation, and investment strategies. Previous studies are contradictory about macroeconomic determinants of housing price and largely focused on one or two areas using point prediction. This study aims to develop data-driven models to accurately predict future housing market trends in different markets. This work studied five different metropolitan areas representing different market trends and compared three-time lagging situations: no lag, 6-month lag, and 12-month lag. Linear regression (LR), random forest (RF), and artificial neural network (ANN) were employed to model the real estate price using datasets with S&P/Case-Shiller home price index and 12 demographic and macroeconomic features, such as gross domestic product (GDP), resident population, personal income, etc. in five metropolitan areas: Boston, Dallas, New York, Chicago, and San Francisco. The data from March 2005 to December 2018 were collected from the Federal Reserve Bank, FBI, and Freddie Mac. In the original data, some factors are monthly, some quarterly, and some yearly. Thus, two methods to compensate missing values, backfill or interpolation, were compared. The models were evaluated by accuracy, mean absolute error, and root mean square error. The LR and ANN models outperformed the RF model due to RF’s inherent limitations. Both ANN and LR methods generated predictive models with high accuracy ( > 95%). It was found that personal income, GDP, population, and measures of debt consistently appeared as the most important factors. It also showed that technique to compensate missing values in the dataset and implementation of time lag can have a significant influence on the model performance and require further investigation. The best performing models varied for each area, but the backfilled 12-month lag LR models and the interpolated no lag ANN models showed the best stable performance overall, with accuracies > 95% for each city. This study reveals the influence of input variables in different markets. It also provides evidence to support future studies to identify the optimal time lag and data imputing methods for establishing accurate predictive models.

Keywords: linear regression, random forest, artificial neural network, real estate price prediction

Procedia PDF Downloads 80

Authors: Mohd Sabri Minhat, Nurul Adilla Mohd Subha

Abstract:

The PUSPATI TRIGA Reactor (RTP), Malaysia reached its first criticality on June 28, 1982, with power capacity 1MW thermal. The Feedback Control Algorithm (FCA) which is conventional Proportional-Integral (PI) controller, was used for present power control method to control fission process in RTP. It is important to ensure the core power always stable and follows load tracking within acceptable steady-state error and minimum settling time to reach steady-state power. At this time, the system could be considered not well-posed with power tracking performance. However, there is still potential to improve current performance by developing next generation of a novel design nuclear core power control. In this paper, the dual-mode predictions which are proposed in modelling Optimal Model Predictive Control (OMPC), is presented in a state-space model to control the core power. The model for core power control was based on mathematical models of the reactor core, OMPC, and control rods selection algorithm. The mathematical models of the reactor core were based on neutronic models, thermal hydraulic models, and reactivity models. The dual-mode prediction in OMPC for transient and terminal modes was based on the implementation of a Linear Quadratic Regulator (LQR) in designing the core power control. The combination of dual-mode prediction and Lyapunov which deal with summations in cost function over an infinite horizon is intended to eliminate some of the fundamental weaknesses related to MPC. This paper shows the behaviour of OMPC to deal with tracking, regulation problem, disturbance rejection and caters for parameter uncertainty. The comparison of both tracking and regulating performance is analysed between the conventional controller and OMPC by numerical simulations. In conclusion, the proposed OMPC has shown significant performance in load tracking and regulating core power for nuclear reactor with guarantee stabilising in the closed-loop.

Keywords: core power control, dual-mode prediction, load tracking, optimal model predictive control

Procedia PDF Downloads 139

Authors: Zhaoan Wang

Abstract:

Research in predictive maintenance modeling has improved in the recent years to predict failures and needed maintenance with high accuracy, saving cost and improving manufacturing efficiency. However, classic prediction models provide little valuable insight towards the most important features contributing to the failure. By analyzing and quantifying feature importance in predictive maintenance models, cost saving can be optimized based on business goals. First, multiple classifiers are evaluated with cross-validation to predict the multi-class of failures. Second, predictive performance with features provided by different feature selection algorithms are further analyzed. Third, features selected by different algorithms are ranked and combined based on their predictive power. Finally, linear explainer SHAP (SHapley Additive exPlanations) is applied to interpret classifier behavior and provide further insight towards the specific roles of features in both local predictions and global model behavior. The results of the experiments suggest that certain features play dominant roles in predictive models while others have significantly less impact on the overall performance. Moreover, for multi-class prediction of machine failures, the most important features vary with type of machine failures. The results may lead to improved productivity and cost saving by prioritizing sensor deployment, data collection, and data processing of more important features over less importance features.

Keywords: automated supply chain, intelligent manufacturing, predictive maintenance machine learning, feature engineering, model interpretation

Procedia PDF Downloads 106

Authors: Wendu Pang, Yaxin Luo, Junhong Li, Yu Zhao, Danni Cheng, Yufang Rao, Minzi Mao, Ke Qiu, Yijun Dong, Fei Chen, Jun Liu, Jian Zou, Haiyang Wang, Wei Xu, Jianjun Ren

Abstract:

We aimed to compare the prognostic prediction value of positive lymph node number (PLNN) to the American Joint Committee on Cancer (AJCC) tumor, lymph node, and metastasis (TNM) staging system for patients with hypopharyngeal squamous cell carcinoma (HPSCC). A total of 826 patients with HPSCC from the Surveillance, Epidemiology, and End Results database (2004–2015) were identified and split into two independent cohorts: training (n=461) and validation (n=365). Univariate and multivariate Cox regression analyses were used to evaluate the prognostic effects of PLNN in patients with HPSCC. We further applied six Cox regression models to compare the survival predictive values of the PLNN and AJCC TNM staging system. PLNN showed a significant association with overall survival (OS) and cancer-specific survival (CSS) (P < 0.001) in both univariate and multivariable analyses, and was divided into three groups (PLNN 0, PLNN 1-5, and PLNN>5). In the training cohort, multivariate analysis revealed that the increased PLNN of HPSCC gave rise to significantly poor OS and CSS after adjusting for age, sex, tumor size, and cancer stage; this trend was also verified by the validation cohort. Additionally, the survival model incorporating a composite of PLNN and TNM classification (C-index, 0.705, 0.734) performed better than the PLNN and AJCC TNM models. PLNN can serve as a powerful survival predictor for patients with HPSCC and is a surrogate supplement for cancer staging systems.

Keywords: hypopharyngeal squamous cell carcinoma, positive lymph nodes number, prognosis, prediction models, survival predictive values

Procedia PDF Downloads 117

Authors: Alan Wan

Abstract:

In recent years, the body of literature on frequentist model averaging in statistics has grown significantly. Most of this work focuses on models with different mean structures but leaves out the variance consideration. In this paper, we consider a regression model with multiplicative heteroscedasticity and develop a model averaging method that combines maximum likelihood estimators of unknown parameters in both the mean and variance functions of the model. Our weight choice criterion is based on a minimisation of a plug-in estimator of the model average estimator's squared prediction risk. We prove that the new estimator possesses an asymptotic optimality property. Our investigation of finite-sample performance by simulations demonstrates that the new estimator frequently exhibits very favourable properties compared to some existing heteroscedasticity-robust model average estimators. The model averaging method hedges against the selection of very bad models and serves as a remedy to variance function misspecification, which often discourages practitioners from modeling heteroscedasticity altogether. The proposed model average estimator is applied to the analysis of two real data sets.

Keywords: heteroscedasticity-robust, model averaging, multiplicative heteroscedasticity, plug-in, squared prediction risk

Procedia PDF Downloads 341

Authors: Vijay Pal Singh, Yogesh Chandra Kotiyal

Abstract:

Structures are a combination of various load carrying members which transfer the loads to the foundation from the superstructure safely. At the design stage, the loading of the structure is defined and appropriate material choices are made based upon their properties, mainly related to strength. The strength of materials kept on reducing with time because of many factors like environmental exposure and deformation caused by unpredictable external loads. Hence, to predict the strength of materials used in structures, various techniques are used. Among these techniques, Non-Destructive Techniques (NDT) are the one that can be used to predict the strength without damaging the structure. In the present study, the compressive strength of concrete has been predicted using Artificial Neural Network (ANN). The predicted strength was compared with the experimentally obtained actual compressive strength of concrete and equations were developed for different models. A good co-relation has been obtained between the predicted strength by these models and experimental values. Further, the co-relation has been developed using two NDT techniques for prediction of strength by regression analysis. It was found that the percentage error has been reduced between the predicted strength by using combined techniques in place of single techniques.

Keywords: rebound, ultra-sonic pulse, penetration, ANN, NDT, regression

Procedia PDF Downloads 401

Authors: Mahdi Alshamasin, Riad Al-Kasasbeh, Nikolay Korenevskiy

Abstract:

We apply mathematical models for the interaction of the internal and biologically active points of meridian structures. Amongst the diseases for which reflex diagnostics are effective are those of the stomach disease. It is shown that use of fuzzy logic decision-making yields good results for the prediction and early diagnosis of gastrointestinal tract diseases, depending on the reaction energy of biologically active points (acupuncture points). It is shown that good results for the prediction and early diagnosis of diseases from the reaction energy of biologically active points (acupuncture points) are obtained by using fuzzy logic decision-making.

Keywords: acupuncture points, fuzzy logic, diagnostically important points (DIP), confidence factors, membership functions, stomach diseases

Procedia PDF Downloads 441