• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• 2019.11a
• /
• pp.266-266
• /
• 2019

Marine accidents are increasing year by year, and various accidents occur such as engine failure, collision, stranding, and fire. These marine accidents present a risk of large casualties. It is important to prevent accidents beforehand. In this study, we propose a modeling to predict the occurrence of marine accidents by applying the Markov Chain Process that can predict the future based on past data. Applying the proposed modeling, the probability of future marine accidents was calculated and compared with the actual frequency. Through this, a probabilistic model was proposed to prepare a prediction system for marine accidents, and it is expected to contribute to predicting various marine accidents.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.17 no.4
• /
• pp.375-382
• /
• 2011

There is no system to obtain the basic data and proceed data and user input interface is complex, thus there are some limitation to utilize the oil spill prediction model. It is difficult to build the scientific response strategy in order to respond oil spill accident rapidly because it is impossible to operate the oil spill prediction model any time. In this study, the optimum operational system for oil spil prediction model has been developed considering the present system. External real time data has been linked because of impossibility of building all basic data and minimum database has been build in this study. Through this data system, real time oil spill prediction model can be utilized. And the user interface has been designed to reduce the error of the interface between user and model and the output interface has been proposed to analyze the result of modeling at multidimensional aspect. While the system for oil spill prediction model as the result of this study has some uncertainties because of depending on external data, the thing that we can predict oil spill using operate the model rapidly as soon as the accident occurred can be meaning in the response field.

• Journal of Korean Society of Transportation
• /
• v.25 no.3
• /
• pp.123-135
• /
• 2007

The objective of this study is to analyze the relationship between traffic accidents occurring at trumpet interchange ramps according to accident type as well as the relevant factors that led to the traffic accidents, such as geometric design elements and traffic volumes. In the process of analysis of the distribution of traffic accidents, negative binomial distribution was selected as the most appropriate model. Negative binomial regression models were developed for total trumpet interchange ramps, direct ramps, loop ramps and semi-direct ramps based on the negative binomial distribution. Based upon several statistical diagnostics of the difference between observed accidents and predicted accidents with four previously developed models, the fit proved to be reasonable. Understanding of statistically significant variables in the developed model will enable designers to increase efficiency in terms of road operations and the development of traffic accident prevention policies in accordance with road design features.

• Design & Manufacturing
• /
• v.14 no.4
• /
• pp.40-45
• /
• 2020

In a vehicle-to-vehicle accident, the impact posture, braking status, final stopping position, collision point and collision speed are important factors for accident reconstruction. In particular, the speed of collision is the most important issue. In this study, the collision speed and the final stopping position in the tunnel were estimated using PC-CRASH, a vehicle crash analysis program used for traffic accident analysis, and the final stopping position of the simulation and the final stopping position of the traffic accident report were compared. When the Pride speed was 0km/h or 30km/h and the Sorento speed was 100m/h, the simulation results and reports matched the final stopping positions and posture of the two vehicles. As a result of the simulation, it can be estimated that Pride was collided in an almost stationary state.

• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.6
• /
• pp.11-20
• /
• 2023

The cost of medical treatment for motor vehicle accidents is increasing every year. In this study, we created a model to predict long-term hospitalization(more than 18 days) among minor patients, which is the main item of increasing traffic accident medical expenses, using five algorithms such as decision tree, and analyzed the factors affecting long-term hospitalization. As a result, the accuracy of the prediction models ranged from 91.377 to 91.451, and there was no significant difference between each model, but the random forest and XGBoost models had the highest accuracy of 91.451. There were significant differences between models in the importance of explanatory variables, such as hospital location, name of disease, and type of hospital, between the long-stay and non-long-stay groups. Model validation was tested by comparing the average accuracy of each model cross-validated(10 times) on the training data with the accuracy of the validation data. To test of the explanatory variables, the chi-square test was used for categorical variables.

• Journal of Korean Society for Quality Management
• /
• v.24 no.1
• /
• pp.10-19
• /
• 1996

In order to build a model to predict accidents in a complicated man-machine sytem, human errors and mechanical reliability can be viewed as the most important factors. Such factors are explicitly included in a generic model. Another point to keep in mind is that the model should be constructed so that the data in a type of accident can be utilized to predict other types of accidents. Based on such a generic prediction model, we analyze the effects of system reliability. When we improve the system reliability, in other words, when there are changes in model parameters, the predicted time to next accidents should be modified influencing the effects of system reliability improvements. We apply Bayesian approach and finds the formula to explain how a change on the machine reliability or human error probability influences the time to next accident.

• Nuclear Engineering and Technology
• /
• v.52 no.11
• /
• pp.2565-2571
• /
• 2020

Background: Understanding the behaviour of nuclear fuel claddings by conducting burst test on single cladding tube under simulated loss-of-coolant accident conditions and developing theoretical cum empirical predictive computer codes have been the focus of several investigations. The developed burst criterion (a) assumes symmetrical deformation of cladding tube in contrast to experimental observation (b) interpolates the properties of Zircaloy-4 cladding in mixed α+β phase (c) does not account for azimuthal temperature variations. In order to overcome all these drawbacks of burst criterion, it is reasoned that artificial intelligence technique may be a better option to predict the burst parameters. Methods: Artificial neural network models based on feedforward backpropagation algorithm with logsig transfer function are developed. Results: Neural network architecture of 2-4-4-3, that is model with two hidden layers having four nodes in each layer is found to be the most suitable. The mean, maximum, and minimum prediction errors for this optimised model are 0.82%, 19.62%, and 0.004%, respectively. Conclusion: The burst stress, burst temperature, and burst strain obtained from burst criterion have average deviation of 19%, 12%, and 53% respectively whereas the developed neural network model predicted these parameters with average deviation of 6%, 2%, and 8%, respectively.

• Nuclear Engineering and Technology
• /
• v.55 no.12
• /
• pp.4607-4616
• /
• 2023

We implement machine learning regression models to predict peak pressures of primary and secondary systems, a major safety concern in Loss Of Condenser Vacuum (LOCV) accident. We selected the Multi-dimensional Analysis of Reactor Safety-KINS standard (MARS-KS) code to analyze the LOCV accident, and the reference plant is the Korean Optimized Power Reactor 1000MWe (OPR1000). eXtreme Gradient Boosting (XGBoost) is selected as a machine learning tool. The MARS-KS code is used to generate LOCV accident data and the data is applied to train the machine learning model. Hyperparameter optimization is performed using a simulated annealing. The randomly generated combination of initial conditions within the operating range is put into the input of the XGBoost model to predict the peak pressure. These initial conditions that cause peak pressure with MARS-KS generate the results. After such a process, the error between the predicted value and the code output is calculated. Uncertainty about the machine learning model is also calculated to verify the model accuracy. The machine learning model presented in this paper successfully identifies a combination of initial conditions that produce a more conservative peak pressure than the values calculated with existing methodologies.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.11 no.6
• /
• pp.703-710
• /
• 2018

Although various efforts have been made every year to reduce electric fire accidents such as accident analysis and inspection for electric fire accidents, there is no effective countermeasure due to lack of effective decision support system and existing cumulative data utilization method. The purpose of this study is to develop an algorithm for predicting electric fire based on data such as electric safety inspection data, electric fire accident information, building information, and weather information. Through the pre-processing of collected data for each institution such as Korea Electrical Safety Corporation, Meteorological Administration, Ministry of Land, Infrastructure, and Transport, Fire Defense Headquarters, convergence, analysis, modeling, and verification process, we derive the factors influencing electric fire and develop prediction models. The results showed insulation resistance value, humidity, wind speed, building deterioration(aging), floor space ratio, building coverage ratio and building use. The accuracy of prediction model using random forest algorithm was 74.7%.

• Nuclear Engineering and Technology
• /
• v.52 no.12
• /
• pp.2836-2846
• /
• 2020

During severe nuclear power plant (NPP) accidents, a H₂/CO mixture can be generated in the reactor pressure vessel by core degradation and in the containment as well by molten corium-concrete interaction. In spite of its importance, a state-of-the-art methodology predicting H₂/CO combustion risk relies predominantly on empirical correlations. It is therefore necessary to develop a proper methodology for flammability evaluation of H₂/CO mixtures at ex-vessel phases characterized by three factors: CO concentration, high temperature, and diluents. The developed methodology adopted Le Chatelier's law and a calculated non-adiabatic flame temperature model. The methodology allows the consideration of the individual effect of the heat transfer characteristics of hydrogen and carbon monoxide on low flammability limit prediction. The accuracy of the developed model was verified using experimental data relevant to ex-vessel phase conditions. With the developed model, the prediction accuracy was improved substantially such that the maximum relative prediction error was approximately 25% while the existing methodology showed a 76% error. The developed methodology is expected to be applicable for flammability evaluation in chemical as well as NPP industries.