• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제15권8호
• /
• pp.3086-3101
• /
• 2021

To supply precise marketing and differentiated service for the electric power service department, it is very important to predict the customers with high sensitivity of electric power failure. To solve this problem, we propose a novel grouped 𝑙_1/2 sparsity constrained logistic regression method for sensitivity assessment of electric power failure. Different from the 𝑙₁ norm and k-support norm, the proposed grouped 𝑙_1/2 sparsity constrained logistic regression method simultaneously imposes the inter-class information and tighter approximation to the nonconvex 𝑙₀ sparsity to exploit multiple correlated attributions for prediction. Firstly, the attributes or factors for predicting the customer sensitivity of power failure are selected from customer sheets, such as customer information, electric consuming information, electrical bill, 95598 work sheet, power failure events, etc. Secondly, all these samples with attributes are clustered into several categories, and samples in the same category are assumed to be sharing similar properties. Then, 𝑙_1/2 norm constrained logistic regression model is built to predict the customer's sensitivity of power failure. Alternating direction of multipliers (ADMM) algorithm is finally employed to solve the problem by splitting it into several sub-problems effectively. Experimental results on power electrical dataset with about one million customer data from a province validate that the proposed method has a good prediction accuracy.

• Earthquakes and Structures
• /
• 제27권3호
• /
• pp.177-189
• /
• 2024

Accurately predicting the failure modes of reinforced concrete (RC) columns is essential for structural design and assessment. In this study, the challenges of imbalanced datasets and complex feature selection in machine learning (ML) methods were addressed through an optimized ML approach. By combining feature selection and oversampling techniques, the prediction of seismic failure modes in rectangular RC columns was improved. Two feature selection methods were used to identify six input parameters. To tackle class imbalance, the Borderline-SMOTE1 algorithm was employed, enhancing the learning capabilities of the models for minority classes. Eight ML algorithms were trained and fine-tuned using k-fold shuffle split cross-validation and grid search. The results showed that the artificial neural network model achieved 96.77% accuracy, while k-nearest neighbor, support vector machine, and random forest models each achieved 95.16% accuracy. The balanced dataset led to significant improvements, particularly in predicting the flexure-shear failure mode, with accuracy increasing by 6%, recall by 8%, and F1 scores by 7%. The use of the Borderline-SMOTE1 algorithm significantly improved the recognition of samples at failure mode boundaries, enhancing the classification performance of models like k-nearest neighbor and decision tree, which are highly sensitive to data distribution and decision boundaries. This method effectively addressed class imbalance and selected relevant features without requiring complex simulations like traditional methods, proving applicable for discerning failure modes in various concrete members under seismic action.

• 정보과학회 논문지
• /
• 제42권12호
• /
• pp.1551-1560
• /
• 2015

소프트웨어 신뢰도 성장 모델은 소프트웨어 신뢰도를 정량적으로 평가하기 위해서 사용되며 고장 데이터를 사용해서 소프트웨어 출시일 또는 추가 테스트 노력을 결정하기 위해서도 사용된다. 특정 소프트웨어 신뢰도 성장 모델을 모든 소프트웨어에 사용할 수 없기 때문에 평가 대상 소프트웨어에 가장 잘 맞는 소프트웨어 신뢰도 성장 모델을 선택하는 것이 중요한 이슈가 되었다. 기존 소프트웨어 신뢰도 성장 모델 선택 방법은 수집된 고장 데이터에 대한 소프트웨어 신뢰도 성장 모델의 적합도만을 평가하며 앞으로 발생할 고장 예측의 정확도는 고려하지 않는다. 이 논문에서는 고장 데이터의 트렌드와 고장 예측능력을 반영한 소프트웨어 신뢰도 성장 모델 선택 방법을 제안한다. 연구의 타당성을 보이기 위하여 실험을 통해서 기존 소프트웨어 신뢰도 성장 모델 선택 방법의 문제점을 확인하고 이 논문에서 제안하는 소프트웨어 신뢰도 성장 모델 선택 방법을 사용하면 기존 방법에 비해 더 정확한 고장 예측을 하는 신뢰도 모델을 선택할 수 있음을 보인다.

• 한국신뢰성학회지:신뢰성응용연구
• /
• 제17권4호
• /
• pp.296-304
• /
• 2017

Purpose: Recent weapon systems in defense have increased the complexity and importance of software when developing multifunctional equipment. In this study, we analyze the accuracy of the proposed software reliability model when applied to weapon systems. Methods: Determine the similarity between software reliability analysis results (prediction/estimation) utilizing data from developing weapon systems and system failures data during operation of weapon systems. Results: In case of a software reliability prediction model, the predicted failure rate was higher than the actual failure rate, and the estimation model was consistent with actual failure history data. Conclusion: The software prediction model needs to adjust the variables that are appropriate for the domestic weapon system environment. As the reliability of software is increasingly important in the defense industry, continuous efforts are needed to ensure accurate reliability analysis in the development of weapon systems.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2002년도 춘계학술대회 논문집
• /
• pp.873-876
• /
• 2002

This paper presents the effect of internal corrosion, external corrosion, material properties, operation condition, earthquake, traffic load and design thickness in pipeline on the failure prediction using a failure probability model. A nonlinear corrosion is used to represent the loss of pipe wall thickness with time. The effects of environmental, operational, and design random variables such as a pipe diameter, earthquake, fluid pressure, a corrosion rate, a material yield stress and a pipe thickness on the failure probability are systematically investigated using a failure probability model for the corrosion pipeline.

• 한국신뢰성학회:학술대회논문집
• /
• 한국신뢰성학회 2002년도 정기학술대회
• /
• pp.403-410
• /
• 2002

This paper presents the effect of internal corrosion, external corrosion, material properties, operation condition, earthquake, traffic load and design thickness in pipeline on the failure prediction using a failure probability model. A nonlinear corrosion is used to represent the loss of pipe wall thickness with time. The effects of environmental, operational, and design random variables such as a pipe diameter, earthquake, fluid pressure, a corrosion rate, a material yield stress and a pipe thickness on the failure probability are systematically investigated using a failure probability model for the corrosion pipeline.

• Journal of Mechanical Science and Technology
• /
• 제20권4호
• /
• pp.473-481
• /
• 2006

Predicting fatigue life by numerical methods was almost impossible in the field of rubber materials. One of the reasons is that there is not obvious fracture criteria caused by nonstandardization of material and excessively various way of mixing process. But, tearing energy as fracture factor can be applied to a rubber-like material regardless of different types of fillers, relative to other fracture factors and the crack growth process of rubber could be considered as the whole fatigue failure process by the existence of potential defects in industrial rubber components. This characteristic of fatigue failure could make it possible to predict the fatigue life of rubber components in theoretical way. FESEM photographs of the surface of industrial rubber components were analyzed for verifying the existence and distribution of potential defects. For the prediction of fatigue life, theoretical way of evaluating tearing energy for the general shape of test-piece was proposed. Also, algebraic expression for the prediction of fatigue life was derived from the rough cut growth rate equation and verified by comparing with experimental fatigue lives of dumbbell fatigue specimen in various loading condition.

• 상하수도학회지
• /
• 제30권2호
• /
• pp.155-165
• /
• 2016

Pipe Deterioration Prediction (PDP) and Pipe Failure Risk Prediction (PFRP) models were developed in an attempt to predict the deterioration and failure risk in water mains using fuzzy technique and the markov process. These two models were used to determine the priority in repair and replacement, by predicting the deterioration degree, deterioration rate, failure possibility and remaining life in a study sample comprising 32 water mains. From an analysis approach based on conservative risk with a medium policy risk, the remaining life for 30 of the 32 water mains was less than 5 years for 2 mains (7%), 5-10 years for 8 (27%), 10-15 years for 7 (23%), 15-20 years for 5 (17%), 20-25 years for 5 (17%), and 25 years or more for 2 (7%).

• Journal of the Korean Data and Information Science Society
• /
• 제8권1호
• /
• pp.21-30
• /
• 1997

가속수명시험에서 강한충격수준에서 부품들의 고장시간이 관측되고 가속화된 고장시간을 토대로 정상충격수준에서 부품들의 성능을 조사한다. 본 논문은 지수수명분포에서 중도절단된 가속수명자료를 이용하여 고장률의 사전분포의 평균을 알 때, 정상조건하에서 하나의 미래 관찰치의 예측문제를 사전분포의 모수에 대하여 적률추정량을 이용하는 경험적 베이즈 접근방법을 적용시켜 경험적 베이즈 예측분포와 예측구간에 대하여 연구하였다.

• Nuclear Engineering and Technology
• /
• 제56권9호
• /
• pp.3740-3749
• /
• 2024

To reduce the losses caused by aging failure of insulation gate bipolar transistor (IGBT), which is the core components of nuclear power plant rod position indicating and rod control (RPC) system. It is necessary to conduct studies on its life prediction. The selection of IGBT failure characteristic parameters in existing research relies heavily on failure principles and expert experience. Moreover, the analysis and learning of time-domain degradation data have not been fully conducted, resulting in low prediction efficiency as the monotonicity, time correlation, and poor anti-interference ability of extracted degradation features. This paper utilizes the advantages of the stacked denoising autoencoder(SDAE) network in adaptive feature extraction and denoising capabilities to perform adaptive feature extraction on IGBT time-domain degradation data; establishes a long-short-term memory (LSTM) prediction model, and optimizes the learning rate, number of nodes in the hidden layer, and number of hidden layers using the Gray Wolf Optimization (GWO) algorithm; conducts verification experiments on the IGBT accelerated aging dataset provided by NASA PCoE Research Center, and selects performance evaluation indicators to compare and analyze the prediction results of the SDAE-LSTM model, PSOLSTM model, and BP model. The results show that the SDAE-LSTM model can achieve more accurate and stable IGBT life prediction.