Search

In the domain of intelligent surveillance for public safety, rapid anomaly detection in crowded environments is essential. This study presents an approach to crowd behaviour analysis by measuring crowd energy changes. Image pixels are modeled as particles, and optical flow techniques are used to extract velocity vectors and directions. To mitigate the noise, occlusions, and lighting challenges of optical flow, the system incorporates pixel motion estimation across frames, improving temporal coherence for smoother motion. Image grey entropy and Otsu’s segmentation are employed to separate foreground from background, enabling detailed energy distribution analysis. Abnormal crowd activity is detected by observing sudden changes in motion intensity. Evaluation on the UMN dataset shows that the proposed method achieves an accuracy of 96.87% in anomaly detection, outperforming other conventional techniques. These results highlight the improved accuracy and efficiency of the method in detecting anomalous crowd behaviour in complex environments. [ABSTRACT FROM AUTHOR]

To tackle the defects of self-supervised learning anomaly detection methods for wireless sensor network (WSN) need to address the problems of single negative sample types and lack of diversity, as well as insufficient extraction of spatiotemporal features from multimodal data of wireless sensor network nodes. To address these challenges, a wireless sensor network anomaly node detection method that combines contrastive learning and reconstruction mechanisms was proposed. Firstly, this method provided sufficient positive and negative example information representation for the reconstruction model by using contrastive learning methods, and combined with generative adversarial network (GAN) to generate negative examples with diverse characteristics. Secondly, a dual layer spatiotemporal feature extraction module based on multi-head attention and graph neural network was designed. Through a series of comparative experiments on actual public datasets and their experimental results, it is shown that the method designed has better accuracy and recall compared to traditional anomaly detection methods and recent graph neural network methods.

Simple Summary: Efficient breeding of meat ducks using three-dimensional and multi-layer cages is a novel approach being actively explored in China. In this process, timely and accurate detection of abnormal situations among ducks is crucial for optimizing and refining the cage-rearing system, and ensuring animal health and welfare. This study focused on the overturned and dead status of cage-reared ducks using YOLOv8 as the basic network. By introducing GAM and Wise-IoU loss functions, we proposed an abnormal-situation recognition method for cage-reared ducks based on YOLOv8-ACRD. Building on this, we refined the identification of key body parts of cage-reared ducks, focusing on six key points: head, beak, chest, tail, left foot, and right foot. This resulted in the development of an abnormal posture estimation model for cage-reared ducks, based on HRNet-48. Furthermore, through multiple tests and comparative verification experiments, it was confirmed that the proposed method exhibited high detection accuracy, generalization ability, and robust comprehensive performance. The method proposed in this study for perceiving abnormal situations in cage-reared ducks not only provides foundational information for the progress and improvement of the meat duck cage-reared system but also offers technological references for the intelligent breeding of other cage-reared poultry. Overturning and death are common abnormalities in cage-reared ducks. To achieve timely and accurate detection, this study focused on 10-day-old cage-reared ducks, which are prone to these conditions, and established prior data on such situations. Using the original YOLOv8 as the base network, multiple GAM attention mechanisms were embedded into the feature fusion part (neck) to enhance the network's focus on the abnormal regions in images of cage-reared ducks. Additionally, the Wise-IoU loss function replaced the CIoU loss function by employing a dynamic non-monotonic focusing mechanism to balance the data samples and mitigate excessive penalties from geometric parameters in the model. The image brightness was adjusted by factors of 0.85 and 1.25, and mainstream object-detection algorithms were adopted to test and compare the generalization and performance of the proposed method. Based on six key points around the head, beak, chest, tail, left foot, and right foot of cage-reared ducks, the body structure of the abnormal ducks was refined. Accurate estimation of the overturning and dead postures was achieved using the HRNet-48. The results demonstrated that the proposed method accurately recognized these states, achieving a mean Average Precision (mAP) value of 0.924, which was 1.65% higher than that of the original YOLOv8. The method effectively addressed the recognition interference caused by lighting differences, and exhibited an excellent generalization ability and comprehensive detection performance. Furthermore, the proposed abnormal cage-reared duck pose-estimation model achieved an Object Key point Similarity (OKS) value of 0.921, with a single-frame processing time of 0.528 s, accurately detecting multiple key points of the abnormal cage-reared duck bodies and generating correct posture expressions. [ABSTRACT FROM AUTHOR]

Copyright of Chinese Journal of Liquid Crystal & Displays is the property of Chinese Journal of Liquid Crystal & Displays and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The structural health monitoring data of the Research Institute of Highway Ministry of Transportation Track (RIOHTRACK) are huge and complex, including a large amount of dynamic high-frequency sensor data of road underground structures. However, detecting anomalies in the overall distribution of the whole loading cycle data is difficult for traditional numerical data analysis methods. This study proposes an anomaly detection method that visualizes numerical values and designs a deep convolutional neural network DCNN6 for image classification to achieve anomaly detection of large-scale dynamic high-frequency sensor data. After training, the detection rate of DCNN6 for abnormal data reached 92.3% for the validation set. Compared with Residual Neural Network (ResNet50) and GhostNet, the detection accuracy of the method proposed in this study increased by 69% and 4%, respectively, reaching 97%, and the detection speeds were also faster by 5 s/epoch and 4 s/epoch, respectively. Therefore, the proposed method can accurately and quickly detect the abnormality of the dynamic high-frequency sensor data of underground structures, which can provide data support for quickly discovering that the vehicle deviates from the preset trajectory, rectifying the driver's driving deviation, analyzing the force of the whole road area, and grasping the evolution law of the rut. [ABSTRACT FROM AUTHOR]

Net Primary Productivity (NPP) is a critical metric for assessing terrestrial carbon sequestration and ecosystem health. While advancements in NPP modeling have enabled estimation at various scales, hidden anomalies within NPP time series necessitate further investigation to understand the driving forces. This study focuses on Shandong Province, China, generating a high-resolution (250 m) monthly NPP product for 2000–2019 using the Carnegie–Ames–Stanford Approach (CASA) model, integrated with satellite remote sensing and ground observations. We employed the Seasonal Mann–Kendall (SMK) Test and the Breaks For Additive Season and Trend (BFAST) algorithm to differentiate between gradual declines and abrupt losses, respectively. Beyond analyzing land use and land cover (LULC) transitions, we utilized Random Forest models to elucidate the influence of environmental factors on NPP changes. The findings revealed a significant overall increase in annual NPP across the study area, with a moderate average of 503.45 gC/(m2·a) during 2000–2019. Although 69.67% of the total area displayed a substantial monotonic increase, 3.89% of the area experienced abrupt NPP losses, and 8.43% exhibited gradual declines. Our analysis identified LULC transitions, primarily driven by urban expansion, as being responsible for 55% of the abrupt loss areas and 33% of the gradual decline areas. Random Forest models effectively explained the remaining areas, revealing that the magnitude of abrupt losses and the intensity of gradual declines were driven by a complex interplay of factors. These factors varied across vegetation types and change types, with explanatory variables related to vegetation status and climatic factors—particularly precipitation—having the most prominent influence on NPP changes. The study suggests that intensified land use and extreme climatic events have led to NPP diminishment in Shandong Province. Nevertheless, the prominent positive vegetation growth trends observed in some areas highlight the potential for NPP enhancement and carbon sequestration through targeted management strategies. [ABSTRACT FROM AUTHOR]

The widespread adoption of video surveillance systems in public security and network security domains has underscored the importance of video anomaly detection as a pivotal research area. To enhance the precision and robustness of anomaly detection, this manuscript introduces an innovative method for video anomaly detection. The approach begins with the application of multi-scale feature extraction technology to capture visual information across varying scales in video data. Leveraging the Spatial Pyramid Convolution (SPC) module as the cornerstone for multi-scale feature learning, the study addresses the impact of scale variations, thereby augmenting the model’s detection capabilities across different scales. Furthermore, a Weakly Supervised Data Augmentation Network (WSDAN) module is incorporated to facilitate attention-guided data augmentation, enhancing the richness of input images. These augmented images undergo training with the U-Net network to elevate detection accuracy. Additionally, the integration of the improved Convolutional Block Attention Module (CBAM) into the base U-Net architecture enables end-to-end training. CBAM dynamically adjusts feature map weights, allowing the model to concentrate on anomaly relevant regions in the video while suppressing interference from non-anomalous areas. To assess anomalies, the paper employs the Peak Signal-to-Noise Ratio (PSNR) between predicted and original frames, normalizing PSNR values for anomaly identification. The proposed method is then evaluated using publicly available datasets CUHK Avenue and UCSD Ped2, with results visually presented. Experimental findings showcase Area Under the Receiver-Operating Characteristic Curve (AUC) values of 86.2% and 97.9% for these datasets, surpassing comparative methods and confirming the effectiveness and superiority of the proposed approach. [ABSTRACT FROM AUTHOR]

Under the X-ray scanning, mobile phone explosive modified at the battery is stealthy, which increases the difficulty of security inspections to detect prohibited phones. This critical issue has not received enough attention. In this paper, we contribute the first modified mobile phone X-ray image benchmark for object recognition, named MPXray. MPXray focuses on the detection of prohibited items that are imbalanced and fine-grained. To deal with such anomaly detection task where few abnormal samples are obtained, we propose a few-shot prohibited phone detection (FSPPD) model based on contrastive learning. FSPPD uses an unsupervised sampling module(USM) to obtain anchors that are more representative of the data distribution, so as to construct balanced input for contrastive learning. For handling hard-to-classify caused by fine-grained samples, an anchor-wise contrastive loss(AW-CL) is designed to supervise models speed up the proximity between intra-class samples and separation between between-class samples. FSPPD is more suitable for applications where electronic products need to be checked individually. We evaluate our model on MPXray, from both the classification perspective and anomaly detection perspective. Experimental results show that our model achieves better recall for modified mobile phones. Additionally, we verify the generalization ability of the proposed model on the CIFAR10 dataset. Compared with widely used algorithms, our model achieves certain superiority in recall metrics. [ABSTRACT FROM AUTHOR]

Structural health monitoring is an effective method to evaluate the safety status of dams. Measurement error is an important factor which affects the accuracy of monitoring data modeling. Processing the abnormal monitoring data before data analysis is a necessary step to ensure the reliability of the analysis. In this paper, we proposed a method to process the abnormal dam displacement monitoring data on the basis of matrix manipulation and Cuckoo Search algorithm. We first generate a scatter plot of the monitoring data and exported the matrix of the image. The scatter plot of monitoring data includes isolate outliers, clusters of outliers, and clusters of normal points. The gray scales of isolated outliers are reduced using Gaussian blur. Then, the isolated outliers are eliminated using Ostu binarization. We then use the Cuckoo Search algorithm to distinguish the clusters of outliers and clusters of normal points to identify the process line. To evaluate the performance of the proposed data processing method, we also fitted the data processed by the proposed method and by the commonly used 3- σ method using a regression model, respectively. Results indicate that the proposed method has a better performance in abnormal detection compared with the 3- σ method. [ABSTRACT FROM AUTHOR]

Safety is the foundation of urban sustainable development. The urban construction and operation process involves a large amount of multidimensional time series data. By detecting anomalies in these multidimensional time subsequences (MTSs), decision support can be provided for early warning of urban construction and operation risks. Considering the complexity of urban infrastructure, there is an urgent need for fast and accurate anomaly detection. This paper proposes a real-time anomaly detection algorithm based on improved distance measurement (RADIM). RADIM retains the relationships between dimensions in multidimensional subsequences, using an Extended Frobenius Norm with Local Weights (EFN_lw) and a Euclidean distance based on multidimensional data (ED_mv) to measure the similarity of MTSs. Moreover, a threshold update mechanism based on First-order Mean Difference (TMFD) is designed to detect real-time anomalies by assessing deviations. This method has been applied to tunnel construction. According to comparative experiments, RADIM exhibits better adaptability, real-time performance, and accuracy in risk warning of tunnel boring machines and construction status. [ABSTRACT FROM AUTHOR]

Introduction: Intelligent power distribution systems are vital in the modern power industry, tasked with managing power distribution efficiently. These systems, however, encounter challenges in anomaly detection, hampered by the complexity of data and limitations in model generalization. Methods: This study developed a Transformer-GAN model that combines Transformer architectures with GAN technology, efficiently processing complex data and enhancing anomaly detection. This model's self-attention and generative capabilities allow for superior adaptability and robustness against dynamic data patterns and unknown anomalies. Results: The Transformer-GAN model demonstrated remarkable efficacy across multiple datasets, significantly outperforming traditional anomaly detection methods. Key highlights include achieving up to 95.18% accuracy and notably high recall and F1 scores across diverse power distribution scenarios. Its exceptional performance is further underscored by achieving the highest AUC of 96.64%, evidencing its superior ability to discern between normal and anomalous patterns, thereby reinforcing the model's advantage in enhancing the security and stability of smart power systems. Discussion: The success of the Transformer-GAN model not only boosts the stability and security of smart power distribution systems but also finds potential applications in industrial automation and the Internet of Things. This research signifies a pivotal step in integrating artificial intelligence into the power sector, promising to advance the reliability and intelligent evolution of future power systems. [ABSTRACT FROM AUTHOR]

Falls among the elderly are a significant public health issue, resulting in about 684,000 deaths annually. Such incidents often lead to severe consequences including fractures, contusions, and cranial injuries, immensely affecting the quality of life and independence of the elderly. Existing fall detection methods using cameras and wearable sensors face challenges such as privacy concerns, blind spots in vision and being troublesome to wear. In this paper, we propose PIFall, a Pressure Insole-Based Fall Detection System for the Elderly, utilizing the ResNet3D algorithm. Initially, we design and fabricate a pair of insoles equipped with low-cost resistive films to measure plantar pressure, arranging 5 × 9 pressure sensors on each insole. Furthermore, we present a fall detection method that combines ResNet(2+1)D with an insole-based sensor matrix, utilizing time-series 'stress videos' derived from pressure map data as input. Lastly, we collect data on 12 different actions from five subjects, including fall risk activities specifically designed to be easily confused with actual falls. The system achieves an overall accuracy of 91% in detecting falls and 94% in identifying specific fall actions. Additionally, feedback is gathered from eight elderly individuals using a structured questionnaire to assess user experience and satisfaction with the pressure insoles. [ABSTRACT FROM AUTHOR]

The increasingly large server clusters of state grid corporation generate a large amount of production operation data, and real-time analysis of the massive monitoring data generated by various devices and systems has become a new challenge in power IT operation and maintenance work. As a key technology of intelligent grid information operation and maintenance work, anomaly detection technology can effectively detect operation and maintenance faults and provide timely alarms to avoid damage to sensitive equipment. Currently, some traditional anomaly detection methods have few types of anomalies and low precision, resulting in delayed fault detection. To address this challenge, this article proposes a multi-dimensional time series anomaly detection method based on capsule networks, NNCapsNet. Firstly, the unsupervised algorithm is applied in combination with expert knowledge to preprocess and label the performance monitoring data of grid marketing business application servers. Secondly, the capsule network is introduced for classification and anomaly detection. Experimental results obtained through five-fold cross-validation show that NNCapsNet achieves an average classification accuracy of 91.21% on a dataset containing 15 types of anomalies. At the same time, compared with four benchmark models on the dataset containing 20 000 monitoring data, NNCapsNet achieves good results in key evaluation indicators. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Xi'an Jiaotong University is the property of Editorial Office of Journal of Xi'an Jiaotong University and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

The present era is characterized by the interconnection, communication, connectivity, and data exchange of Internet of Things (IoT) devices. However, current systems often neglect to incorporate security protocols for IoT devices in the energy sector, the Internet of Medical Things, smart homes, and other areas. This poses a significant challenge in IoT networks as these devices possess constrained resources, making them vulnerable to attacks. Attackers can exploit these vulnerabilities to gain unauthorized access and retrieve sensitive information or data from the targeted devices. This paper presents a machine learning-driven intrusion detection system to tackle these issues, aiming to devise a system for early identification of such attacks. The system is tested using the WUSTL and UNSW-NB18 datasets to detect Man-in-the-Middle (MITM) and Botnet attacks. The developed system selects the optimal features from both datasets using Mutual Information (MI) and chi-square feature selection. It applies the Synthetic Minority Oversampling Technique (SMOTE) resampling method to resample the attack class and the Random Under Sampling method to resample the normal class. This study utilizes TabNet, Support Vector Machine (SVM), and Random Forest (RF) for both datasets. The performance is then compared with the proposed Ensemble Weighted Voting (EWV) classifier. The experimental results show that the proposed method PSO-EWV on the WUSTL dataset achieves 99.958% and 99.992% F-scores on the UNSW 2018 dataset for MITM attack and Botnet attack classification with MI feature selection. The experimental findings conclude that this method effectively detects attacks within an intrusion detection system.

Introduction: Intelligent power distribution systems are vital in the modern power industry, tasked with managing power distribution efficiently. These systems, however, encounter challenges in anomaly detection, hampered by the complexity of data and limitations in model generalization.Methods: This study developed a Transformer-GAN model that combines Transformer architectures with GAN technology, efficiently processing complex data and enhancing anomaly detection. This model’s self-attention and generative capabilities allow for superior adaptability and robustness against dynamic data patterns and unknown anomalies.Results: The Transformer-GAN model demonstrated remarkable efficacy across multiple datasets, significantly outperforming traditional anomaly detection methods. Key highlights include achieving up to 95.18% accuracy and notably high recall and F1 scores across diverse power distribution scenarios. Its exceptional performance is further underscored by achieving the highest AUC of 96.64%, evidencing its superior ability to discern between normal and anomalous patterns, thereby reinforcing the model’s advantage in enhancing the security and stability of smart power systems.Discussion: The success of the Transformer-GAN model not only boosts the stability and security of smart power distribution systems but also finds potential applications in industrial automation and the Internet of Things. This research signifies a pivotal step in integrating artificial intelligence into the power sector, promising to advance the reliability and intelligent evolution of future power systems.

Advancements in AI have elevated speech recognition, with convolutional neural networks (CNNs) proving effective in processing spectrogram-transformed speech signals. CNNs, with lower parameters and higher accuracy compared to traditional models, are particularly efficient for deployment on storage-limited embedded devices. Artificial neural networks excel in predicting inputs within their expected output range but struggle with anomalies. This is usually harmful to a speech recognition system. In this paper, the neural network classifier for speech recognition is trained with a "negative branch" method, incorporating directional regularization with out-of-distribution training data, allowing it to maintain a high confidence score to the input within distribution while expressing a low confidence score to the anomaly input. It can enhance the performance of anomaly detection of the classifier, addressing issues like misclassifying the speech command that is out of the distribution. The result of the experiment suggests that the accuracy of the CNN model will not be affected by the regularization of the "negative branch", and the performance of abnormal detection will be improved as the number of kernels of the convolutional layer increases. [ABSTRACT FROM AUTHOR]

Due to its widespread application in the field of public security, anomaly detection in crowd scenes has recently become a hot topic. Some deep learning-based methods led to significant accomplishments in this field. Nevertheless, due to the scarcity of data and the misclassification of queries which most of them suffer to some extent from a sudden and infrequent overfitting. Though, we tried to solve the above problems, understand the long video streams and establish an accurate and reliable security system in order to improve its performance in detecting anomalies. We also referred to the hash technique, which has proven to be the most efficient method used when researching about large-scale image recovery. Thus, this article offers a smart video anomaly detection solution. In this paper, we combine the advantages of both deep hashing and deep auto-encoders to show that tracking changes in deep hash components across time and can be used to detect local anomalies. More precisely, we start with a new technique to minimize the mass of input data and information in order to decrease the time of calculation using a new dynamic frame skipping technique. Then, we propose to measure local anomalies by combining semantic with low-level optical flows to balance the performance and perceptibility. The experimental results illustrate that the proposed methods surpass these baselines for the detection and localization of anomalies. [ABSTRACT FROM AUTHOR]

Detecting anomalous crowd behavioral patterns from videos is an important task in video surveillance and maintaining public safety. In this work, we propose a novel architecture to detect anomalous patterns of crowd movements via graph networks. We represent individuals as nodes and individual movements with respect to other people as the node-edge relationship of an evolving graph network. We then extract the motion information of individuals using optical flow between video frames and represent their motion patterns using graph edge weights. In particular, we detect the anomalies in crowded videos by modeling pedestrian movements as graphs and then by identifying the network bottlenecks through a max-flow/min-cut pedestrian flow optimization scheme (MFMCPOS). The experiment demonstrates that the proposed framework achieves superior detection performance compared to other recently published state-of-the-art methods. Considering that our proposed approach has relatively low computational complexity and can be used in real-time environments, which is crucial for present day video analytics for automated surveillance. [ABSTRACT FROM AUTHOR]

Within the Shuo Huang Railway Company (Suning, China) the long-term evolution for railways (LTE-R) network carries core wireless communication services for trains. The communication performance of LTE-R cells directly affects the operational safety of the trains. Therefore, this paper proposes a novel detection method for LTE-R cells with degraded communication performance. Considering that the number of LTE-R cells with degraded communication performance and that of normal cells are extremely imbalanced and that the communication performance indicator data for each cell are sequence data, we propose a feature extraction neural network structure for imbalanced sequences, based on shapelet transformation and a convolutional neural network (CNN). Then, to train the network, we set the optimization objective based on the Fisher criterion. Finally, using a two-stage training method, we obtain a neural network model that can distinguish LTE-R cells with degraded communication performance from normal cells at the feature level. Experiments on a real-world dataset show that the proposed method can realize the accurate detection of LTE-R cells with degraded communication performance and has high practical application value. [ABSTRACT FROM AUTHOR]

The wind turbine power curve model is critical to a wind turbine's power prediction and performance analysis. However, abnormal data in the training set decrease the prediction accuracy of trained models. This paper proposes a sample average approach-based method to construct an interval model of a wind turbine, which increases robustness against abnormal data and further improves the model accuracy. We compare our proposed methods with the traditional neural network-based and Bayesian neural network-based models in experimental data-based validations. Our model shows better performance in both accuracy and computational time. [ABSTRACT FROM AUTHOR]

Named entity recognition (NER) model can recognize the normal entities, but cannot provide any hints for the missing entity or the entity in the incorrect location, which cannot meet the extensive requirements on the text entities in the information processing and the archiving analysis field. Combined with the specific context characteristics of the entity anomaly recognition, a method for the entity location anomaly and the entity absence anomaly detection (NER-EAD) integrating GCNN (gated convolutional neural network) and GRU (gated recurrent unit) and its training data construction method are proposed, which is based on the structure of the named entity recognition model with the pre-trained language model. GCNN extracts more character context features to improve the model performance in identifying abnormal entities. The method integrates the semantic feature output of the convolutional neural network and the recurrent neural network to comprehensively extract the features of normal entities and the abnormal entities. Experiments show that NER-EAD reaches average F1 of 90.56%, 85.56% and 80.92% in the normal entity recognition, the entity location anomaly detection and the entity absence anomaly detection, respectively, surpassing the existing named entity recognition model architecture. Additionally, the ablation experiment proves the semantic feature extraction ability of the fusion network of GCNN and GRU.

This study proposes a methodology for diagnosing the degree of performance degradation of the adsorbent in pressure swing adsorption (PSA) plants using a one-dimensional simulator and a time-series deep learning algorithm. First, a 1D PSA simulator was developed using mathematical models and validated with previously published experimental data. The behavior change of the PSA plant according to the performance degradation was trained using a deep learning algorithm based on the developed simulator. The model combines the 1D convolutional neural network and long-short-term memory (LSTM) network. The prediction of the degradation degree of the internal adsorbent was then presented using a pretrained neural network. The developed methodology demonstrates a mean squared error lower than 10−6 when predicting the degree of adsorbent degradation from the adsorption-bed-temperature time-series profiles with an example. The methodology can be used to predictive maintenance strategy by identifying PSA performance degradation in real time without stopping operation. [ABSTRACT FROM AUTHOR]

Featured Application: This method addresses the problem of predicting anomaly data in software runtime. The proposed method utilizes multiple time windows to choose models from multiple classes of anomaly method detectors and fuses the anomaly results in order to save the process of anomaly detection model selection. In practice, the technique can make timely predictions and alerts for anomalous operation. The detection of anomalies in software systems has become increasingly crucial in recent years due to their impact on overall software quality. However, existing integrated anomaly detectors usually combine the results of multiple detectors in a clustering manner and do not consider the changes in data anomalies in the time dimension. This paper investigates the limitations of existing anomaly detection methods and proposes an improved integrated anomaly detection approach based on time windows and a voting mechanism. By utilizing multiple time windows, the proposed method overcomes the challenges of cumulative anomalies and achieves enhanced performance in capturing anomalies that accumulate gradually over time. Additionally, two hybrid models are introduced, based on accuracy and sensitivity, respectively, to optimize performance metrics such as AUC, precision, recall, and F1-score. The proposed method demonstrates remarkable performance, achieving either the highest or only a marginal 3% lower performance compared to the optimal model. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Harbin Institute of Technology. Social Sciences Edition / Haerbin Gongye Daxue Xuebao. Shehui Kexue Ban is the property of Harbin Institute of Technology and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Since the leakage of sodium in an SFR (sodium-cooled fast reactor) causes an explosion upon reaction with air and water, sodium leakages represent an important safety issue. In this study, a novel technique for improving the reliability of sodium leakage detection applying DDVR (dynamic data validation and reconciliation) is proposed and verified to resolve this technical issue. DDVR is an approach that aims to improve the accuracy of a target system in a dynamic state by minimizing random errors, such as from the uncertainty of instruments and the surrounding environment, and by eliminating gross errors, such as instrument failure, miscalibration, or aging, using the spatial redundancy of measurements in a physical model and the reliability information of the instruments. DDVR also makes it possible to estimate the state of unmeasured points. To validate this approach for supporting sodium leakage detection, this study applies experimental data from a sodium leakage detection experiment performed by the Korea Atomic Energy Research Institute. The validation results show that the reliability of sodium leakage detection is improved by cooperation between DDVR and hardware measurements. Based on these findings, technology integrating software and hardware approaches is suggested to improve the reliability of sodium leakage detection by presenting the expected true state of the system.

The wind turbine power curve model is critical to a wind turbine’s power prediction and performance analysis. However, abnormal data in the training set decrease the prediction accuracy of trained models. This paper proposes a sample average approach-based method to construct an interval model of a wind turbine, which increases robustness against abnormal data and further improves the model accuracy. We compare our proposed methods with the traditional neural network-based and Bayesian neural network-based models in experimental data-based validations. Our model shows better performance in both accuracy and computational time.

Network anomaly detection faces unique challenges from dynamic traffic, including large data volume, few attributes, and human factors that influence it, making it difficult to identify typical behavioral characteristics. To address this, we propose using Sketch-based Profile Evolution (SPE) to detect network traffic anomalies. Firstly, the Traffic Graph (TG) of the network terminal is generated using Sketch to identify abnormal data flow positions. Next, the Convolutional Neural Network and Long Short-Term Memory Network (CNN-LSTM) are used to develop traffic behavior profiles, which are then continuously updated using Evolution to detect behavior pattern changes in real-time data streams. SPE allows for direct processing of raw traffic datasets and continuous detection of constantly updated data streams. In experiments using real network traffic datasets, the SPE algorithm was found to be far more efficient and accurate than PCA and Basic Evolution for outlier detection. It is important to note that the value of φ can affect the results of anomaly detection. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Frontiers of Computer Science & Technology is the property of Beijing Journal of Computer Engineering & Applications Journal Co Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Abnormal detection of wind turbine converter (WT) is one of the key technologies to ensure long-term stable operation and safe power generation of WT. The number of normal samples in the SCADA data of WT converter operation is much larger than the number of abnormal samples. In order to solve the problem of low abnormal data and low recognition rate of WTs, we propose a sample enhancement method for WT abnormality detection based on an improved conditional Wasserstein generative adversarial network. Since the anomaly samples of WT converters are few and difficult to obtain, the CWGANGP oversampling method is constructed to increase the anomaly samples in the WT converter dataset. The method adds additional category labels to the inputs of the generative and discriminative models of the generative adversarial network, constrains the generative model to generate few types of anomalous samples, and enhances the generative model's ability to generate few types of anomalous samples, enabling data generation in a prescribed direction. The smooth continuous Wasserstein distance is used instead of JS divergence as a distance metric to measure the probability distribution of real and generated data in the conditional generative response network and reduce pattern collapse. The gradient constraint is added to the CWGANGP model to enhance the convergence of the WGAN model, so that the generative model can synthesize minority class anomalous samples more effectively and accurately under the condition of unbalanced sample data categories. The quality of anomalous sample generation is also improved. Finally, the anomaly detection is made on the actual operating variator dataset for the unbalanced dataset and the dataset after reaching Nash equilibrium. The experimental results show that the method used in this paper has lower MAR and FAR in WT converter anomaly detection compared with other oversampling data balance optimization methods such as SMOTE, RandomOverSampler, GAN, etc. The method can be well implemented for anomaly detection of large wind turbines and can be better applied in WT intelligent systems. [ABSTRACT FROM AUTHOR]

Copyright of Cyber Security & Data Governance is the property of Editorial Office of Information Technology & Network Security and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

As the growth of social network scale, so do challenges to the existing anomaly detection algorithms. Therefore, this paper proposed an anomaly detection method based on graph modularity clustering（GMC＿AD）, which could be applied to solve the problem of low detection efficiency caused by network size and complexity. Based on analyzing the network topology structure, the GMC＿AD method improved the efficiency of events detection by weighting mechanism on abnormal nodes and modularity clustering algorithm. The GMC＿AD processes could be described as follow: a)Since designing a quantization strategy for node evolution in the network, GMC＿AD get the set of abnormal nodes by recognizing nodes with abnormal evolutionary behaviors. b)The method used a modularity clustering algorithm to reduce the network size. c)During the calculation of network fluctuation value, it introduced a weighting mechanism for taking the influence of abnormal nodes into consideration, after that, the GMC＿AD method detected the abnormality by the changes of network fluctuation value. On real social network datasets VAST, EU＿E-mail and ENRON, the GMC＿AD method accurately detected the abnormal periods. The event detection sensibility of GMC＿AD method was increased by 50%～82% meanwhile the run-time efficiency increased by 30%～70%. The GMC＿AD method enhances not only the accuracy and sensitivity but also the efficiency of anomaly detections. [ABSTRACT FROM AUTHOR]