• Smart Structures and Systems
• /
• v.33 no.4
• /
• pp.301-311
• /
• 2024

This paper proposes a deep convolutional long short-term memory (ConvLSTM)-based crack growth prediction technique for predictive maintenance of structures. Since cracks are one of the critical damage types in a structure, their regular inspection has been mandatory for structural safety and serviceability. To effectively establish the structural maintenance plan using the inspection results, crack propagation or growth prediction is essential. However, conventional crack prediction techniques based on mathematical models are not typically suitable for tracking complex nonlinear crack propagation mechanism on civil structures under harsh environmental conditions. To address the technical issue, a field data-driven crack growth prediction technique using ConvLSTM is newly proposed in this study. The proposed technique consists of the four steps: (1) time-series crack image acquisition, (2) target image stabilization, (3) deep learning-based crack detection and quantification and (4) crack growth prediction. The performance of the proposed technique is experimentally validated using a concrete mock-up specimen by applying step-wise bending loads to generate crack growth. The validation test results reveal the prediction accuracy of 94% on average compared with the ground truth obtained by field measurement.

• Journal of Ocean Engineering and Technology
• /
• v.9 no.2
• /
• pp.112-122
• /
• 1995

In this work, prediction of fatigue life and fatigue crack growth are studied. 4th order polynominal function is presented to describe the crack growth behaviors from artifical pit of SM45C steel. Crack growth curves obtained from 4th order polyminal growth equations are in good agreement with experimental data The crack growth behaviors at arbitrary stress levels and investigated by the concept of elastic-plastic fracture mechanics using ${\Delta}J$. Fatigue life prediction are carried out by numerical integral method. Prediction lives obtained by proposed method in this study, is in good agreement with the experimental ones. Life prediction results calculated by using of ${\Delta}J$ better than those of ${\Delta}K$.

• Agribusiness and Information Management
• /
• v.6 no.1
• /
• pp.1-11
• /
• 2014

The purpose of this study is to develop the growth prediction model that can predict growth and development information influencing on the production of citrus fruits. The growth model was developed to predict the floral leaf ratio, number of fruit sets, fruit width, and overweight fruits depending on the main period of growth and development by considering the weather factors because the fruit production is influenced by weather depending on the growth and development period. To predict the outdoor-grown citrus fruit production, the investigation result for the standard farms is used as the basic data; in this study, we also understood that the influence of weather factors on the citrus fruit production based on the data from 2004 to 2013 of the outdoor-grown citrus fruit observation report in which the standard farms were targeted by the Agricultural Research Service and suggested the growth and development information prediction model with the weather information as an independent variable to build the observation model. The growth and development model for outdoor-grown citrus fruits was assumed by using the Ordinary Least Square method (OLS), and the developed growth prediction model can make a prediction in advance with the weather factors prior to the observation investigation for the citrus fruit production. To predict the growth and development information of the production of citrus fruits having a great ripple effect as a representative crop in Jeju agriculture, the prediction result regarding the production applying the weather factors depending on growth and development period could be applied usefully.

• Advances in nano research
• /
• v.3 no.3
• /
• pp.143-168
• /
• 2015

Initiation of crack and its growth simulation requires accurate model of traction - separation law. Accurate modeling of traction-separation law remains always a great challenge. Atomistic simulations based prediction has great potential in arriving at accurate traction-separation law. The present paper is aimed at establishing a method to address the above problem. A method for traction-separation law prediction via utilizing atomistic simulations data has been proposed. In this direction, firstly, a simpler approach of common neighbor analysis (CNA) for the prediction of crack growth has been proposed and results have been compared with previously used approach of threshold potential energy. Next, a scheme for prediction of crack speed has been demonstrated based on the stable crack growth criteria. Also, an algorithm has been proposed that utilizes a variable relaxation time period for the computation of crack growth, accurate stress behavior, and traction-separation atomistic law. An understanding has been established for the generation of smoother traction-separation law (including the effect of free surface) from a huge amount of raw atomistic data. A new curve fit has also been proposed for predicting traction-separation data generated from the molecular dynamics simulations. The proposed traction-separation law has also been compared with the polynomial and exponential model used earlier for the prediction of traction-separation law for the bulk materials.

• Korea Trade Review
• /
• v.46 no.6
• /
• pp.191-209
• /
• 2021

In this paper, various time series analysis models and machine learning models are presented for long-term prediction of export growth rate, and the prediction performance is compared and reviewed by RMSE and MAE. Export growth rate is one of the major economic indicators to evaluate the economic status. And It is also used to predict economic forecast. The export growth rate may have a negative (-) value as well as a positive (+) value. Therefore, Instead of using the ReLU function, which is often used for time series prediction of deep learning models, the PReLU function, which can have a negative (-) value as an output value, was used as the activation function of deep learning models. The time series prediction performance of each model for three types of data was compared and reviewed. The forecast data of long-term prediction of export growth rate was deduced by three forecast methods such as a fixed forecast method, a recursive forecast method and a rolling forecast method. As a result of the forecast, the traditional time series analysis model, ARDL, showed excellent performance, but as the time period of learning data increases, the performance of machine learning models including LSTM was relatively improved.

• Communications for Statistical Applications and Methods
• /
• v.17 no.6
• /
• pp.909-916
• /
• 2010

A trend in software reliability engineering is to take into account the coverage growth behavior during testing. A coverage growth function that represents the coverage growth behavior is an essential factor in software reliability models. When multiple competitive coverage growth functions are available, there is a need for a criterion to select the best coverage growth functions. This paper proposes a selection criterion based on the prediction error. The conditional coverage growth function is introduced for predicting future coverage growth. Then the sum of the squares of the prediction error is defined and used for selecting the best coverage growth function.

• Journal of Korean Society for Quality Management
• /
• v.47 no.1
• /
• pp.187-198
• /
• 2019

Purpose: In this study, a reliability prediction based reliability growth management is suggested especially for the early development phase of a system and the case study of surveillance system is given. Methods: The proposed reliability prediction based reliability growth management procedures consists of 7 Steps. In Step 1, the stages for reliability growth management are classified according to the major design changes. From Step 2 to Step 5, system reliability is predicted based on reliability structures and the predicted reliabilities of subsystems (Level 2) and modules (Level 3). At each stage, by comparing the predicted system reliability with that of the previous stage, the reliability growth of the system is checked in Step 6. In Step 7, when the predicted value of sustem reliability does not satisfy the reliability goal, some design alternatives are considered and suggested to improve the system reliability. Results: The proposed reliability prediction based reliability growth management can be an efficient alternative for managing reliability growth of a system in its early development phase. The case study shows that it is applicable to weapon system such as a surveillance system. Conclusion: In this study, the procedures for a reliability prediction based reliability growth management are proposed to satisfy the reliability goal of the system efficiently. And it is expected that the use of the proposed procedures would reduce, in the test and evaluation phase, the number of corrective actions and its cost as well.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.16 no.8
• /
• pp.1536-1544
• /
• 1992

A relatively simple prediction method is proposed for initially semicircular surface crack growth under axial loading. The method takes into account the difference in surface crack closure behavior at the depth point and at the surface intersection point, and also the relationship of crack closure for surface crack and through-thickness crack. The prediction method provides conservative estimation for fatigue life within factor of two, and the predicted crack geometry variations agree well with the observed results. As a result, the prediction method proposed here is considered to be useful for engineering application.

• Transactions of the Korean Society of Mechanical Engineers
• /
• v.14 no.6
• /
• pp.1583-1591
• /
• 1990

Life prediction and residual life prediction of structures of machines are one of the most strongly world wide needed problems as requirement in the stage of slowly developing economy which comes after rapidly and highly developing stage. For the purpose of statistical life prediction, fatigue test was conducted under the 4 stress levels, and for each stress level, about 20 specimens are used. The statistical properties of crack growth parameter m and C in the fatigue crack growth law of da/dN=C(.DELTA.K)sup m/, and the relationship between m and C, and the statistical distribution pattern of fatigue crack growth rate can be obtained by experimental results.

• Korean System Dynamics Review
• /
• v.11 no.3
• /
• pp.5-31
• /
• 2010

The purpose of this study is to develop a system dynamics model for growth prediction of low birth weight infants(LBWIs) based on nutrition. This growth prediction model consists of 9 modules; body weight, height, carbohydrate, protein, lipid, micronutrient, water, activity and energy module. The results of the model simulation match well with the percentiles of weights and heights of the Korean infants, also with the growth records of 55 LBWIs, under 37 weeks of gestational age, whose weights are appropriate for their gestational age. This model can be used to understand the current growth mode of LBWIs, predict the future growth of LBWIs, and be utilized as a tool for controlling the nutrient intake for the optimal growth of LBWIs in actual practice.