International Journal of System Assurance Engineering and Management最新文献

After the pandemic, the whole world is transforming digital, due to the increased usage of handheld devices like smartphones and due to the evolution of the internet. All the transactions are becoming online. The security at end devices is an important issue to everyone. We believe that the data in transit is more secure, but in reality this is not true. The data are in the hands of bad actors for malicious activities. Android ransomware is one of the most widely distributed assaults throughout the world. It is a type of virus that prevents users from accessing the operating system and encrypts the essential data saved on their device. This work focuses on thorough assessment and detection of android ransomware application using machine learning methods. After a thorough analysis of existing mechanisms of android ransomware detection, we found that the combination of static behaviour with machine learning techniques can detect android ransomware with good accuracy. We have analysed 3572 samples of ransomware applications and 3628 samples of benign applications of various family. For classification, the decision tree, random forest, extra tree classifier, light gradient boosting machine methods are selected from the pool of classifier. The dataset was obtained from Kaggle, which is an open source dataset repository. The suggested model outperforms with a detection accuracy of 98.05%. Based on its best performance, we believe our suggested approach will be useful in ransomware and forensic investigation.

Efficient transportation of fruits and vegetables is crucial for proper storage, handling, and distribution directly influencing their quality, shelf life, and ultimately the price. Maintaining optimal storage conditions during the transport of fruits and vegetables is of utmost importance to preserve their freshness and quality. Therefore, there is a pressing need for a real-time assessment system that can ensure the highest quality and safety of fruits and vegetables throughout the supply chain network. This paper introduces an Internet of Things-enabled sensor network designed to address these challenges. The sensors are strategically deployed within the storage containers that continuously assessing real-time critical environmental parameters, such as temperature, humidity, pH, and air quality. These parameters significantly affect the storage of fruits and vegetables throughout the supply chain network. Furthermore, we have employed machine learning algorithms, such as decision trees, k-nearest neighbors, logistic regression, and Support Vector Machine, to measure performance in terms of accuracy, F1-score, precision, sensitivity, and specificity. The results indicate that the Support Vector Machine algorithm outperforms with the other algorithms with an impressive accuracy of 98.05%. Future research endeavors will focus on optimizing food supply chain loss.

The railway system is a complex technical system-of-systems (SoS). To address the complexity of the railway system, a holistic approach is needed that facilitates the development of an appropriate asset management regime. A systems-of-systems (SoS) approach considers the complex nature of the railway system, comprising interconnected subsystems like rolling stock and infrastructure. Neglecting these interdependencies risks sub-optimization of the overall system performance. Asset management of the railway system utilising a SoS approach ensures the focus of asset management on overall system requirements. The efficiency and effectiveness of the railway system is based on aspects such as availability, reliability, and safety performance. To enhance these aspects, monitoring, and improvement of key performance indicators (KPIs) emphasizing increased capacity and reduced operational costs is essential. The KPIs offer quantifiable parameters for performance optimization. Augmenting asset management through data-driven technologies can improve the efficiency and effectiveness of asset management. However, challenges persist in the implementation of data-driven solutions due to the railway system’s complexity and lack of a holistic perspective. A systematic performance-driven framework with a system-of-systems approach for augmented asset management of railway system provides handrail for the utilisation of data-driven technologies with railway system requirements at the centre while developing an asset management regime. The proposed framework aims to establish a clear relationship between system KPIs, and the performance of sub-systems and components aiding railway organizations in asset management design and implementation. This paper explains the important components of the proposed framework and demonstrates the application the framework for asset management and maintenance planning of high value components in the fleet of railway rolling stock. Adoption of the proposed framework is expected to enhance asset management through development and implementation of data-driven solutions that are aligned with system KPIs, to support asset management decision making.

Depth maps (DMs) are invaluable tools encapsulating scene information in a three-dimensional context. They have a crucial part in reconstructing the spatial layout of a scene, enabling a comprehensive understanding of object geometry. These DMs can originate from either a single image or a combination of multiple images, with the former approach referred to as monocular depth mapping. However, deriving accurate depth maps is a complex and ill-posed problem that often necessitates intricate calibration. Recent advances have turned to deep learning (DL) techniques to address these challenges. In the context of monocular depth estimation, we propose a novel methodology utilizing an Attention U-Net architecture (Attention UNet). By incorporating attention mechanisms, we bolster the network’s ability to extract salient features, particularly along object boundaries. Critically, this enhancement is achieved without introducing additional parameters to the networks, ensuring efficient model training. Our proposed approach is effective in producing high-quality depth maps with notable advantages. By leveraging the Attention UNet architecture, we substantially improve depth map accuracy, reducing the root mean square error (RMSE) by 0.23 on the benchmark NYU V2 dataset, Highlighting its supremacy compared to current state-of-the-art techniques.

Frequency deviation and Tie-Line power flow deviation are major concern due to the continuous load changing condition and the utilization of renewable energy sources in multi microgrid interconnected systems. Therefore, it is important and crucial to maintain the frequency and Tie-line power flow. In this paper, Novel hybrid algorithm combines both Particle Swarm Optimization (PSO) and Grey Wolf Optimization (GWO) driven proportional-integral-derivative (PID) controller and cascade Proportional Integral and Proportional Derivative (PI–PD) controller is suggested to deal with the issues in a proposed multi interconnected microgrid system. At first, the performance of the developed hybrid algorithm driven PID controller is investigated and its performance is compared with individual PSO and GWO driven PID controller. Finally the hybrid algorithm performance is investigated in cascade PI–PD controller and its performance is compared with the PID controller. Integral time multiplied by absolute error (ITAE) is used as the objective function in this work for obtaining optimum parameters of both PID and PI–PD controller. The simulated results show the superiority of the proposed hybrid algorithm (PSO–GWO) driven PI–PD controller compared with the other techniques in settling time, overshoot etc.

The occurrence of a Loss Of Offsite Power (LOOP) event can be a major threat to nuclear safety due to the dependence of auxiliary systems on electrical energy. Probabilistic safety assessments of nuclear power plants require, thus, estimates of the frequencies and durations of such LOOP events. These estimates are usually based on past statistical data, which is not always relevant. Model-based approaches are thus needed. This paper proposes an analytical method to estimate the frequency and duration of switchyard-centered LOOP events, which constitute one of the four main categories of LOOP events. The proposed method is mainly based on the identification of active minimal cut sets, considering the behavior of circuit breakers against faults according to their coordination and selectivity. Adapted versions of the Risk Reduction Worth and Fussel–Vesely importance factors are proposed to evaluate the impact of components on the switchyard-centered LOOP event frequency. Furthermore, uncertainty analysis is developed and performed. Various generic plant connection schemes are used for application. Results demonstrate the applicability of the methodology to estimate the frequency and duration of switchyard-centered LOOP events, and to identify optimal ways to reduce the risk by modifying the switchyard configuration.

Building safety has become a serious and important topic for the development of the construction industry, as well as for the preservation of contractors' and workers' lives and property. With the development and expansion of a sensitive and complex monitoring system for the safety of buildings, allowing accidents to occur is no longer acceptable. Therefore, risk management identifies potential hazards before any operations take place, and the safety system operates based on a planned, organized, and systematic process known as "pre-incident." This plan is based on the analysis-control method. Failure to utilize risk management methods and the acceleration of the construction industry can lead to a decrease in the safety of residents and introduce unpredictable risks. While nowadays risk management is less utilized for project control, contractors face numerous problems after construction. Lack of resources and facilities in this regard can be problematic, but emerging building technologies, which are slowly being identified, can solve and separate most of the industry's safety issues. Therefore, utilizing innovative building technologies not only enhances quality, speed, and cost reduction in construction but also contributes significantly to industrialization and the reduction of risks resulting from deteriorated structures towards building safety. In this study, the extraordinary effects of innovative technologies on building safety have been examined, and the relationship between risk management and innovative technologies has been investigated using a questionnaire. The impacts of all risk management and safety aspects are examined in this research, which ultimately resulted in clarifying the direct and meaningful connection between risk management and safety with modern technologies and determining the necessary corrective measures to improve building safety performance through the use of innovative building technologies.

In the modern world the availability of the machinery for any industry is of utmost importance. It is the right maintenance at right time which keeps these machineries available for their jobs. The primary goal of maintenance is to avoid or mitigate consequences of failure of equipment. There are various types of maintenance schemes available such as breakdown maintenance, preventive maintenance, condition based maintenance etc. Out of all these schemes Reliability Centred Maintenance (RCM) is most recent one and the application of which will enhance the productivity and availability. RCM ensures better system uptime along with understanding of risk involved. RCM has been used in various industries, however, it is very less explored and utilized in marine operations.Hence in the present study maintenance schemes of a marine diesel engine has been considered for optimization using RCM.Failure Modes and Effects Analysis and Fault Tree Analysis (FTA)are some of the basic steps involved in RCM. Due to the scarcity of reliability data particularly in the marine environment some of the components data had to be estimated based on the operating experience. As FTA is based on binary state perspective, assuming the system exist in either functioning or failed state, some of the components (whose performance varies with time and degrades) cannot be modeled using FTA. Hence, in this paper reliability modeling of performance degraded components is dealt with Markov models and the required data is evaluated from condition monitoring techniques. After obtaining the availability of the marine diesel engine, based on the importance ranking, critical components have been obtained for optimizing the maintenance schedules. In this paper genetic algorithm approach has been used for optimization. The results obtained have been compared and new maintenance scheme has been proposed.

Digital technology has increased the spread of fake news, leading to misconceptions, misunderstandings, and economic challenges. Researchers have developed automated techniques to identify false information using various data features, driven by advancements in AI. Most algorithms focus on signals from the news itself and its context, often ignoring user preferences. According to confirmation bias theory, individuals are more likely to spread false information that aligns with their beliefs. Users’ historical and social activities, such as their postings, can help identify fake news and inform their news choices. However, there is limited research on incorporating user preferences in fake news detection. This study introduces a framework based on Graph Neural Networks (GNNs) and natural language models to capture signals from both graph and content perspectives, considering user preferences. We chose GNNs for their ability to model complex relationships in graph-structured data. Specifically, we used the Graph Attention Network due to its ability to weigh the importance of different nodes, enhancing the capture of relevant signals. The framework integrates user preferences by analyzing social activities and news choices. Experimental results on a real-world dataset show our model achieves an accuracy of 98%. Outperforming models that do even consider user preferences. These findings highlight the potential of leveraging user preferences to enhance fake news detection, offering a more robust approach to tackling information pollution.

Early-stage breast cancer detection remains a critical challenge in healthcare, demanding innovative approaches that leverage the power of deep learning and transfer learning techniques. The problem to be investigated involves designing a model capable of extracting meaningful features from mammographic images, maximizing transferability across datasets, and optimizing the trade-off between model complexity and computational efficiency. Existing methods often face limitations in achieving high accuracy, robustness, and efficiency. This research aims to address these challenges by proposing a novel transfer learning approach that combines the strengths of VGG11 and EfficientNet architectures for early-stage breast cancer detection. In the case of technological development, there is never a shortage of opportunities in the field of medical imaging. Cancer patients who have an earlier diagnosis of their disease have a lower probability of passing away from their illness. This research proposed an novel early neural network based on transfer learning names as ‘EARLYNET’ to automate breast cancer prediction. In this research, the new hybrid deep learning model was devised and built for distinguishing benign breast tumors from malignant ones. The trials were carried out on the Breast Histopathology Image dataset, and the model was evaluated using a Mobile net founded on the transfer learning method. In terms of accuracy, this model delivers 91.53% accuracy. Explored how the proposed transfer learning framework can enhance the accuracy and reliability of early-stage breast cancer detection, contributing to advancements in medical image analysis and positively impacting patient outcomes.