Search Results (6)

In nuclear power plants, the loss-of-coolant accident (LOCA) stands out as the most prevalent and consequential incident. Accurate breach size diagnosis is crucial for the mitigation of LOCAs, and identifying the cause of an accident can prevent catastrophic consequences. Traditional methods mostly focus on combining model algorithms and utilize intricate composite model neural network architectures. However, it is crucial to investigate whether greater complexity necessarily leads to better performance. In addition, the consideration of the impact of dataset construction and data preprocessing on model performance is also needed for model building. This paper proposes a framework named DeepLOCA-Lattice to experiment with different preprocessing approaches to fundamental deep learning models for a comprehensive analysis of the diagnosis of LOCA breach size. The DeepLOCA-Lattice involves data preprocessing via the lattice algorithm and equal-interval partitioning and deep-learning-based models, including the multi-layer perceptron (MLP), recurrent neural networks (RNNs), convolutional neural networks (CNNs), and the transformer model in LOCA breach size diagnosis. After conducting rigorous ablation experiments, we have discovered that even rudimentary foundational models can achieve accuracy rates that exceed 90%. This is a significant improvement when compared to the previous models, which yield an accuracy rate of lower than 50%. The results interestingly demonstrate the superior performance and efficacy of the fundamental deep learning model, with an effective dataset construction approach. It elucidates the presence of a complex interplay among diagnostic scales, sliding window size, and sliding stride. Furthermore, our investigation reveals that the model attains its highest accuracy within the discussed range when utilizing a smaller sliding stride size and a longer sliding window length. This study could furnish valuable insights for constructing models for LOCA breach size estimation. Full article

Value at risk is a statistic used to anticipate the largest possible losses over a specific time frame and within some level of confidence, usually 95% or 99%. For risk management and regulators, it offers a solution for trustworthy quantitative risk management tools. VaR has become the most widely used and accepted indicator of downside risk. Today, commercial banks and financial institutions utilize it as a tool to estimate the size and probability of upcoming losses in portfolios and, as a result, to estimate and manage the degree of risk exposure. The goal is to obtain the average number of VaR “failures” or “breaches” (losses that are more than the VaR) as near to the target rate as possible. It is also desired that the losses be evenly distributed as possible. VaR can be modeled in a variety of ways. The simplest method is to estimate volatility based on prior returns according to the assumption that volatility is constant. Otherwise, the volatility process can be modeled using the GARCH model. Machine learning techniques have been used in recent years to carry out stock market forecasts based on historical time series. A machine learning system is often trained on an in-sample dataset, where it can adjust and improve specific hyperparameters in accordance with the underlying metric. The trained model is tested on an out-of-sample dataset. We compared the baselines for the VaR estimation of a day (d) according to different metrics (i) to their respective variants that included stock return forecast information of d and stock return data of the days before d and (ii) to a GARCH model that included return prediction information of d and stock return data of the days before d. Various strategies such as ARIMA and a proposed ensemble of regressors have been employed to predict stock returns. We observed that the versions of the univariate techniques and GARCH integrated with return predictions outperformed the baselines in four different marketplaces. Full article

In the context of life sciences, the rapid biotechnical development leads to the creation of huge amounts of biological data. The use of such data naturally brings concerns on human genetic privacy breaches, which also discourage biological data sharing. Prior studies have investigated the possibility of the privacy issues associated with individuals’ trait data. However, there are few studies on quantitatively analyzing the probability of the privacy risk. In this paper, we fill this void by proposing a scheme for systematically breaching genomic privacy, which is centered on quantifying the probability of the privacy risk of continuous trait data. With well-designed synthetic datasets, our theoretical analysis and experiments lead to several important findings, such as: (i) The size of genetic signatures and the sensitivity (true positive rate) significantly affect the accuracy of re-identification attack. (ii) Both the size of genetic signatures and the minor allele frequency have a significant impact on distinguishing true positive and false positive matching between traits and genetic profiles. (iii) The size of the matching quantitative trait locus dataset has a large impact on the confidence of the privacy risk assessment. Validation with a real dataset shows that our findings can effectively estimate the privacy risks of the continuous trait dataset. Full article

Failures of small dams can pose a serious threat to people and property even if the size of the schemes is relatively low. In many cases, small dams are situated in a cascade along streams, meaning that the failure of the uppermost dam may cause the dams downstream to fail. In this paper, a cascade of three small reservoirs, Lichnov II (14.6 m high), Lichnov III (10 m high), and Pocheň (8.5 m high), is the subject of the dam break analyses carried out via various methods such as empirical formulae, analogy, and hydraulic modeling. The dam-break flood routing was simulated using a shallow water flow hydraulic model. The simulations confirm that the attenuation effect of the peak discharge is governed by the flood volume, slope, and morphology of the floodplain and increases with the distance from the breached dam following an approximately exponential trend. When estimating peak discharge, empirical formulae derived for a single dam break should be applied carefully as they may underestimate the peak outflow by up to 10% in the case of a dam cascade. The attenuation volume of small reservoirs is small when compared to the flood volume, meaning that the attenuation of the peak discharge usually varies between 5–10%. Full article

Floods are the most destructive natural phenomenon, by the total number of casualties, and value of property damage, compared to any other type of natural disaster. However, some of the most destructive flash floods are related to dam breaches or complete collapses, that release the large amounts of water, affecting inhabited areas. Worldwide, numerous dams have almost reached or surpassed the estimated construction life span, and pose an increasing risk to structure stability. Considering their continuous degrading state, increasing rainfall aggressiveness, due to climatic changes, technical error, or even human error, there are numerous, potential causes, for which dams could develop breaches and completely fail. This study aims to portray a comparative perspective of flood impact, with real-life consequences, measured by quantifiable parameters, generated from computer simulations of different breach sizes. These parameters include the total flooded surface, water velocity, maximum water depth, number of affected buildings, etc. The analysis was undergone by means of HEC-RAS based 2D hydraulic modeling and GIS, depending on high-accuracy Lidar terrain data and historical hydrological data. As a case study, Drăcșani Lake with the associated Sulița earthfill embankment dam was chosen, being one of the largest and oldest artificial lakes in Romania. Full article

Landslide-generated impulse waves in dammed reservoirs run up the reservoir banks as well as the upstream dam slope. If large enough, the waves may overtop and even breach the dam and cause flooding of the downstream area with hazardous consequences. Hence, for reservoirs in landslide-prone areas, it is important to provide a means to estimate the potential size of an event triggered by landslides along the reservoir banks. This research deals with landslide-generated waves and the overtopping process over the dam crest in a three-dimensional (3D) physical model test, presenting a case study. The model set-up describes the landslide impacting the reservoir in a perpendicular manner, which is often the case in natural settings. Based on the experimental results, dimensionless empirical relations are derived between the overtopping volume and the governing parameters, namely the slide volume, slide release height, slide impact velocity, still-water depth, and upstream dam face slope. Predictive relations for the overtopping volume are presented as applicable for cases relating to the specific model set-up. Measured overtopping volumes are further compared to a two-dimensional (2D) case reported in the literature. An important feature regarding the overtopping process for the 3D case is the variation in time and space, resulting in an uneven distribution of the volume of water overtopping the dam crest. This observation is made possible by the 3D model set-up, and is of value for dam safety considerations as well as for foundation-related issues, including erosion and scouring. Full article