Pattern Recognition Algorithms Research Articles

Variational Mode Decomposition-Based Processing for Detection of Short-Circuited Turns in Transformers Using Vibration Signals and Machine Learning

Transformers are key elements in electrical systems. Although they are robust machines, different faults can appear due to their inherent operating conditions, e.g., the presence of different electrical and mechanical stresses. Among the different elements that compound a transformer, the winding is one of the most vulnerable parts, where the damage of turn-to-turn short circuits is one of the most studied faults since low-level damage (i.e., a low number of short-circuited turns—SCTs) can lead to the overall fault of the transformer; therefore, early fault detection has become a fundamental task. In this regard, this paper presents a machine learning-based method to diagnose SCTs in the transformer windings by using their vibrational response. In general, the vibration signals are firstly decomposed by means of the variational mode decomposition method, where a comparison with the empirical mode decomposition (EMD) method and the ensemble empirical mode decomposition (EEMD) method is also carried out. Then, entropy, energy, and kurtosis indices are obtained from each decomposition as fault indicators, where both the combination of features and the dimensionality reduction by using the principal component analysis (PCA) method are analyzed for the global effectiveness improvement and the computational burden reduction. Finally, a pattern recognition algorithm based on artificial neural networks (ANNs) is used for automatic fault detection. The obtained results show 100% effectiveness in detecting seven fault conditions, i.e., 0 (healthy), 5, 10, 15, 20, 25, and 30 SCTs.

A Triplet Multimodel Transfer Learning Network for Speech Disorder Screening of Parkinson’s Disease

Deterioration in the quality of a person’s voice and speech is an early sign of Parkinson’s disease (PD). Although a number of computer-based methods have been invested to use patients’ speech for early diagnosis of Parkinson’s disease, they only focus on a fixed pronunciation test, such as the subjects’ monosyllabic pronunciation is analyzed to determine whether they have potential possibility of PD. Moreover, only using traditional speech analysis methods to extract single-view speech features cannot provide a comprehensive feature representation. This paper is dedicated to the study of various pronunciation tests for patients with PD, including the pronunciation of five monosyllabic vowels and a spontaneous dialogue. A triplet multimodel transfer learning network is designed and proposed for identifying subjects with PD in these two groups of tests. First, multisource data extract mel frequency cepstrum coefficient (MFCC) features of speech for preprocessing. Subsequently, a pretrained triplet model represents features from three dimensions as the upstream task of the transfer learning framework. Finally, the pretrained model is reconstructed as a novel model that integrates the triplet model, temporal model, and auxiliary layer as the downstream task, and weights are updated through fine-tuning to identify abnormal speech. Experimental results show that the highest PD detection rates in the two groups of tests are 99% and 90% , respectively, which outperform a large number of internationally popular pattern recognition algorithms and serve as a baseline for other academic researchers in this field.

Dazzling Evaluation of the Impact of a High-Repetition-Rate CO2 Pulsed Laser on Infrared Imaging Systems.

This article utilizes the Canny edge extraction algorithm based on contour curvature and the cross-correlation template matching algorithm to extensively study the impact of a high-repetition-rate CO2 pulsed laser on the target extraction and tracking performance of an infrared imaging detector. It establishes a quantified dazzling pattern for lasers on infrared imaging systems. By conducting laser dazzling and damage experiments, a detailed analysis of the normalized correlation between the target and the dazzling images is performed to quantitatively describe the laser dazzling effects. Simultaneously, an evaluation system, including target distance and laser power evaluation factors, is established to determine the dazzling level and whether the target is recognizable. The research results reveal that the laser power and target position are crucial factors affecting the detection performance of infrared imaging detector systems under laser dazzling. Different laser powers are required to successfully interfere with the recognition algorithm of the infrared imaging detector at different distances. And laser dazzling produces a considerable quantity of false edge information, which seriously affects the performance of the pattern recognition algorithm. In laser damage experiments, the detector experienced functional damage, with a quarter of the image displaying as completely black. The energy density threshold required for the functional damage of the detector is approximately 3 J/cm2. The dazzling assessment conclusions also apply to the evaluation of the damage results. Finally, the proposed evaluation formula aligns with the experimental results, objectively reflecting the actual impact of laser dazzling on the target extraction and the tracking performance of infrared imaging systems. This study provides an in-depth and accurate analysis for understanding the influence of lasers on the performance of infrared imaging detectors.

Modified suppressed relative entropy fuzzy c-means clustering algorithm

The Fuzzy C-means (FCM) algorithm is one of the most widely used algorithms in unsupervised pattern recognition. As the intensity of observation noise increases, FCM tends to produce large center deviations and even overlap clustering problems. The relative entropy fuzzy C-means algorithm (REFCM) adds relative entropy as a regularization function to the fuzzy C-means algorithm, which has a good ability for noise detection and membership assignment to observed values. However, REFCM still tends to generate overlapping clusters as the size of the cluster increases and becomes imbalanced. Moreover, the convergence speed of this algorithm is slow. To solve this problem, modified suppressed relative entropy fuzzy c-means clustering (MSREFCM) is proposed. Specifically, the MSREFCM algorithm improves the convergence speed of the algorithm while maintaining the accuracy and anti-noise capability of the REFCM algorithm by adding a suppression strategy based on the intra-class average distance measurement. In addition, to further improve the clustering performance of MSREFCM for multidimensional imbalanced data, the center overlapping problem and the center offset problem of elliptical data are solved by replacing the Euclidean distance in REFCM with the Mahalanobis distance. Experiments on several synthetic and UCI datasets indicate that MSREFCM can improve the convergence speed and classification performance of the REFCM for spherical and ellipsoidal datasets with imbalanced sizes. In particular, for the Statlog dataset, the running time of MSREFCM is nearly one second less than that of REFCM, and the accuracy of MSREFCM is 0.034 higher than that of REFCM.

A new approach for heart disease detection using Motif transform-based CWT's time-frequency images with DenseNet deep transfer learning methods.

Electrocardiogram (ECG) signals are extensively utilized in the identification and assessment of diverse cardiac conditions, including congestive heart failure (CHF) and cardiac arrhythmias (ARR), which present potential hazards to human health. With the aim of facilitating disease diagnosis and assessment, advanced computer-aided systems are being developed to analyze ECG signals. This study proposes a state-of-the-art ECG data pattern recognition algorithm based on Continuous Wavelet Transform (CWT) as a novel signal preprocessing model. The Motif Transformation (MT) method was devised to diminish the drawbacks and limitations inherent in the CWT, such as the issue of boundary effects, limited localization in time and frequency, and overfitting conditions. This transformation technique facilitates the formation of diverse patterns (motifs) within the signals. The patterns (motifs) are constructed by comparing the amplitudes of each individual sample value in the ECG signals in terms of their largeness and smallness. In the subsequent stage, the obtained one-dimensional signals from the MT transformation were subjected to CWT to obtain scalogram images. In the last stage, the obtained scalogram images were subjected to classification using DenseNET deep transfer learning techniques. The combined approach of MT+CWT+DenseNET yielded an impressive success rate of 99.31 %.

Segmentation and classification techniques used to detect early stroke diagnosis using brain magnetic resonance imaging: a review

<p>Stroke is a leading cause of disability and death worldwide. Early diagnosis and treatment are crucial in reducing the risk of stroke-related complications. Brain magnetic resonance imaging (MRI) is a common diagnostic tool used for stroke evaluation. However, manual interpretation of MRI images can be time-consuming and subjective. Machine learning (ML) algorithms have shown promise in automating and improving stroke diagnosis accuracy. This article focuses on classification and segmentation techniques used to detect early stroke diagnosis using brain magnetic imaging. The diagnosis, treatment, and prognosis of complications and patient outcomes in a number of neurological diseases are currently made possible by ML through pattern recognition algorithms. However, the use of MRI is limited because of MRI plays an important role in diagnosing lumbar disc disease. However, the use of MRI is limited due to its high cost and significant operational and processing time. More importantly, MRI is contraindicated in some patients who are claustrophobic or have pacemakers due to the potential for damage. Recent studies have shown that treatment within six hours of a stroke can save a patient's life. Unfortunately, Malaysia is facing a shortage of neuroradiologists, hampering efforts to treat its growing number of stroke patients.</p>

Development of a community severance index for urban areas in the United States: A case study in New York City

Background and aimsTraffic-related exposures, such as air pollution and noise, have a detrimental impact on human health, especially in urban areas. However, there remains a critical research and knowledge gap in understanding the impact of community severance, a measure of the physical separation imposed by road infrastructure and motorized road traffic, limiting access to goods, services, or social connections, breaking down the social fabric and potentially also adversely impacting health. We aimed to robustly quantify a community severance metric in urban settings exemplified by its characterization in New York City (NYC). MethodsWe used geospatial location data and dimensionality reduction techniques to capture NYC community severance variation. We employed principal component pursuit, a pattern recognition algorithm, combined with factor analysis as a novel method to estimate the Community Severance Index. We used public data for the year 2019 at census block group (CBG) level on road infrastructure, road traffic activity, and pedestrian infrastructure. As a demonstrative application of the Community Severance Index, we investigated the association between community severance and traffic collisions, as a proxy for road safety, in 2019 in NYC at CBG level. ResultsOur data revealed one multidimensional factor related to community severance explaining 74% of the data variation. In adjusted analyses, traffic collisions in general, and specifically those involving pedestrians or cyclists, were nonlinearly associated with an increasing level of Community Severance Index in NYC. ConclusionWe developed a high spatial-resolution Community Severance Index for NYC using data available nationwide, making it feasible for replication in other cities across the United States. Our findings suggest that increases in the Community Severance Index across CBG may be linked to increases in traffic collisions in NYC. The Community Severance Index, which provides a novel traffic-related exposure, may be used to inform equitable urban policies that mitigate health risks and enhance well-being.

Sparrow search mechanism-based effective feature mining algorithm for the broken wire signal detection of prestressed concrete cylinder pipe

To enhance the capability of distributed optical fiber (DOF) monitoring system on real-time broken wire signal detection of prestressed concrete cylinder pipe (PCCP), this research proposes a novel effective parameter mining algorithm by utilizing wrapper theory and sparrow search algorithm (SSA). First, the improvement approach of SSA is investigated by applying the adaptive adjustment strategy, the Gaussian mutation perturbation, and the Tent chaotic perturbation. The subset search method is constructed based on the improved SSA (ISSA). Second, the similarity degrees of vibration signals are characterized by the Euclidean distance and the average linkage distance. On this basis, the subset evaluation criterion is established according to the classification accuracy of hierarchical clustering (HC). Finally, the prototype experiment is conducted in which the phase-sensitive optical time-domain reflectometry (Φ-OTDR) monitoring system is employed to collect vibration signals. Then, the effectiveness and the performance of the developed methodology are validated. The following results are drawn. (1) Benchmark function tests indicate that the convergence performance, the optimization ability, and the operation stability of ISSA are superior to those of SSA and the improved particle swarm optimization (IPSO). (2) The original set that has 122 feature parameters is established, and ISSA-HC, IPSO-HC, and SSA-HC extract 24, 31, and 55 effective parameters, respectively. (3) The recognition accuracies of the 3 parameter subsets are 95.83%, 84.17%, and 80.83%, respectively. The parameter subset mined by ISSA-HC has the best pattern recognition ability. This research provides technical support for the broken wire monitoring of in-service PCCP engineering. Meanwhile, it offers a foundation for further developing suitable pattern recognition algorithms to identify PCCP broken wire signals.

Monitoring the Workflow of Earthquake by Using Machine Learning

This study explores the innovative application of machine learning techniques in monitoring and analyzing the workflow associated with earthquakes. The primary objective is to enhance the accuracy and efficiency of earthquake detection, prediction, and response mechanisms the development of machine learning technology enables more robust real-time earthquake monitoring through automated implementations. However, the application of machine learning to earthquake location problems faces challenges in regions with limited available training data. This paper proposes a machine learning-based approach to monitor seismic workflows efficiently. We apply various Machine Learning algorithms, including deep learning neural networks and decision trees, to process and interpret vast amounts of seismic data. These algorithms were trained on historical earthquake records to identify patterns and anomalies that precede seismic events. Our approach not only focused on the detection of seismic activities but also on the prediction of potential aftershock locations and magnitudes. Earthquake monitoring and response systems play a crucial role in mitigating potential damages and ensuring rapid response to seismic events. Traditional methods rely on sensor networks and manual analysis, often leading to delays in detection and response. The system leverages real-time data from seismic sensors, incorporating advanced machine learning algorithms for data analysis and pattern recognition. By employing deep learning models, such as convolution neural networks (CNNs) and recurrent neural networks (RNNs), the system can detect and classify seismic activities accurately. Additionally, anomaly detection techniques enable the identification of unusual seismic patterns that might indicate impending earthquakes or aftershocks. The system integrates geographical information systems (GIS) to map the seismic activities spatially, aiding in the visualization and understanding of seismic patterns across regions. The machine learning models are trained on historical seismic data to enhance their predictive capabilities and adaptability to varying seismic behaviors. This proposed system offers a real-time monitoring solution that can assist in early earthquake detection, accurate event classification, and timely response coordination. Its ability to continuously learn from incoming data ensures adaptability to changing seismic behaviors, thereby improving overall efficiency and reliability in earthquake monitoring workflows.

Modeling the effect of implementation of artificial intelligence powered image analysis and pattern recognition algorithms in concrete industry

AI-powered image analysis and pattern recognition algorithms (IAPRA) are renowned for their capacity to identify concrete flaws, assess strength characteristics, and anticipate the service life of concrete. However, its execution in a concrete building is challenging due to several unknown aspects. This research aims to evaluate the challenges encountered by AI-powered IAPRA and their influence on the concrete industry's digital transformation success. We conducted a quantitative research methodology to identify impediments and success variables associated with AI-powered picture analysis and pattern recognition algorithms. Structural Equation Modeling (SEM) tests were conducted to determine the critical hurdles associated with AI-powered picture analysis and pattern recognition algorithms. Three reliable and valid formative constructs were identified: complexity and privacy, economic and legal, and technology and integration. The developed model revealed the significance of three reflecting constructs: quality control, predictive maintenance, and enhanced productivity. The practical implications include, addressing the identified challenges related to AI-powered IAPRA is crucial for the concrete industry's digital transformation. By prioritizing quality control, predictive maintenance, and enhanced productivity, stakeholders can optimize concrete processes and outcomes. The major limitation of this study is its reliance on the quantitative research approach, which inherently restricts the data collection to the specific features of the sample under investigation.

Partial discharge pattern recognition algorithm of overhead covered conductors based on feature optimization and bidirectional LSTM‐GRU

AbstractRecognition of partial discharge (PD) patterns is essential for insulation diagnosis of covered conductors in overhead lines. Current research has not sufficiently addressed the complex background noise in real environments, and most detection methods depend primarily on feature engineering or deep learning, suggesting potential for improvement in accuracy and efficiency. This has led the authors to propose a PD pattern recognition algorithm that integrates feature selection and deep learning. This algorithm incorporates the design of a discrete wavelet denoising function specifically tailored to the characteristics of PD for data preprocessing. It employs Bayesian optimization algorithms and light gradient boosting machines for characterizing corona discharge features. Furthermore, it develops multi‐scale clustering features and phase‐resolved features for feature fusion, and constructs insightful features based on the light gradient boosting machine. Finally, a novel deep learning model is formulated, demonstrating exceptional detection performance for early faults in covered conductors. Experimental results show that this algorithm attains an Matthews correlation coefficient score of 0.814, a 13.2% improvement over the baseline algorithm's 0.719, and a speed increase of 39.18%. The final accuracy amounts to 97.85%. This algorithm demonstrates exceptional performance in detecting early insulation faults in conductors.

Monitoring the workflow of Earthquake by using Machine Learning

The main goal is to improve the accuracy and efficiency of earthquake detection, prediction and response through automation and, over time, to improve machine learning techniques for earthquake monitoring. However, the application of machine learning to earthquake location problems faces challenges in areas with limited data. This article presents an engine-based approach to monitor earthquake performance. We use a variety of Machine Learning algorithms, including deep learning neural networks and decision trees, for processing and interpret large amounts of seismic data. These algorithms are trained on historical earthquake records to identify patterns and anomalies prior to seismic events. Our approach focuses not only on detecting seismic events, but also on predicting the location and magnitude of earthquakes. Earthquake monitoring and response systems play an important role in reducing potential losses and providing rapid response to seismic events. Conventional methods rely on sensor networks and manual analysis, often with delays in detection and response. The system receives real-time data from seismic sensors, incorporating advanced machine learning algorithms for data analysis and pattern recognition. The system can detect and describe seismic events using deep learning models such as convolutional neural networks (CNN) and random neural networks (RNN).In addition, anomaly detection techniques allow the identification of unusual seismic patterns that may indicate impending earthquakes or earthquakes. The system integrates geographic information systems (GIS) to comprehensively map seismic activity, helping to visualize and understand seismic patterns in the region. Machine learning models are trained on historical seismic data to improve their predictive capabilities and adapt to different seismic behaviors. This proposed system provides a real-time monitoring solution that can assist in early earthquake detection, accurate event classification, and timely response coordination. The ability to continuously learn from incoming data enables adaptation to change in seismic activity, thereby increasing the efficiency and reliability of earthquake monitoring operations.

Mutation Endmember Library Sparse Mixed Abundance Estimation Model for Glioma Margin Determination with Raman Spectroscopy.

The glioma margin is a region of brain tissue where glioblastoma tissue transitions to normal tissue with varying levels of cancer cell concentration. This article uses Raman spectroscopy to detect the glioma margin, which is a fuzzy and uncertain substance that cannot be accurately identified by conventional pattern recognition algorithms. This article applies abundance estimation to Raman spectral unmixing of glioma marginal tissues for the accurate and real-time determination of the tumor surgical boundary during an operation. This article introduces a novel method: the mutation endmember library sparse mixed abundance estimation model. This method adds different representative Raman spectra to each endmember library to account for its dynamic properties, thus reducing errors from such variations and fully capturing the diversity within the substance. Moreover, it uses group sparse endmember bundle decomposition, where each substance endmember library consists of multiple Raman spectra. Fractionally mixed norms are used to ensure intergroup and intragroup sparsity, eliminate redundant spectra, and enhance the generalization ability of the abundance estimation. This method was compared with conventional abundance estimation methods. The experimental results of 112 human glioma margin tissues demonstrate that this method outperforms other methods in terms of accuracy, stability, and generalization ability. This article demonstrates the potential of miniature Raman spectroscopy as a new approach to in vivo and noninvasively determining intraoperative margin assessment.

Potential Locations of Strong Earthquakes in Bulgaria and the Neighbouring Regions

Information about potential earthquake sources is a key issue for seismic hazard assessment. This study presents the application of a phenomenological approach based on pattern recognition to determine the possible locations of strong earthquakes in the Bulgarian region. The technique assumes the origin of strong earthquakes in morphostructural nodes formed around the intersections of morphostructural lineaments identified by morphostructural zoning. For the territory of the Bulgaria and neighbouring regions, 178 nodes were defined in this work. The CORA-3 pattern recognition algorithm identified 59 seismogenic nodes analysing a set of geophysical and geological node’s characteristics. The identified seismogenic nodes are capable to generate earthquakes with magnitude equal to or greater than 6 and are located at the boundaries between the largest tectonic domains: Rila, Pirin, and Rhodope orogens; the Serbian-Macedonian massif; and in the Stara Planina belt. The set of characteristic features of seismogenic nodes indicates that the vicinity of potential nodes is characterized by a high contrast of neotectonic movements of the Earth’s crust and the presence of deep heterogeneities in the Earth’s crust. About 40% of the recognized nodes are not associated with any earthquakes, while the rest of the recognized seismogenic nodes are characterized by an area with a radius of 25 km where earthquakes are known to occur. Part of these “non active” seismogenic nodes are close to the historical events with magnitudes higher than 5.5 since the magnitude and location of historical events have large uncertainties. Another part of the seismogenic nodes may slightly change the location due to the uncertainties in morphostructural zonation. Other nodes may indicate unknown historical seismicity or paleoearthquakes. Defined M6+ seismogenic nodes can fill the potential gaps in the recorded seismicity on the territory of Bulgaria, thus to improve the seismic hazard assessment of the studied region.

Enhancing Seismic Damage Detection and Assessment in Highway Bridge Systems: A Pattern Recognition Approach with Bayesian Optimization.

Highway bridges stand as paramount elements within transportation infrastructure systems. The ability to ensure swift recovery after extreme events, such as earthquakes, is a fundamental trait of resilient communities. Consequently, expediting the recovery process necessitates near real-time diagnosis of structural damage to provide dependable information. In this study, a data-driven approach for damage detection and assessment is investigated, focusing on bridge columns-the pivotal supporting elements of bridge systems-based on simulations derived from nonlinear time history analysis. This research introduces a set of cumulative intensity-based damage features, whose efficacy is demonstrated through unsupervised learning techniques. Leveraging the support vector machine, a prominent pattern recognition algorithm in supervised learning, alongside Bayesian optimization with a Gaussian process, seismic damage detection and assessment are explored. Encouragingly, the methodology yields high estimation accuracies for both binary outcomes (indicating the presence of damage or the occurrence of collapse) and multi-class classifications (indicating the severity of damage). This breakthrough opens avenues for the practical implementation of on-board sensor computing, enabling near real-time damage detection and assessment in bridge structures.

Research on Guide Star Distribution of Sub-Arcsecond Attitude Determination for Microsatellites Reusing Scientific Cameras

Onboard scientific cameras are reused in attitude determination to meet the sub-arcsecond attitude determination accuracy requirements of microsatellites. This approach does not require an additional payload for microsatellites. It involves reusing high-quality optical lenses from the scientific camera and utilizing the peripheral high-quality imaging areas of its square-shaped detector. Separate detectors are placed within these areas as attitude determination detectors to obtain star patterns for closed-loop attitude determination, thereby achieving high-precision attitude determination for microsatellites. The star patterns obtained using this method may pose specific issues due to the relative positions of stars. Through an analysis of the theoretical model that examines the relationship between attitude determination accuracy and the main influencing factors, it is indicated that guide star distribution is one of the main, complex factors determining attitude determination accuracy. A further simulation analysis was conducted on the specific impact of two guide star distribution characteristics—namely, the coverage of guide stars in the attitude determination areas and the proportion of the average field of view occupied by the guide star triangles to the total field of view of the attitude determination areas—on attitude determination accuracy. This study concludes that when the measurement error of the guide stars is bigger than the attitude determination accuracy requirement for its area configuration, four attitude determination areas should be configured. Four attitude determination areas should be prioritized when the measurement error is equal to or smaller than the attitude determination accuracy requirement, followed by the option to configure three attitude determination areas or two symmetric attitude determination areas. When selecting guide stars for star pattern recognition, the guide stars should cover the attitude determination areas as much as possible, and guide stars with a higher proportion of the average field of view occupied by the guide star triangles to the total field of view should be chosen. Finally, experimental validation was conducted using star patterns from dense star fields and sparse star fields. The research results provide an important reference for the optimization of attitude determination area configuration, navigation star catalog construction, and star pattern recognition algorithm research for microsatellites equipped with scientific cameras.

Implementing Computational Thinking Skills in Socio Scientific Issue (SSI) of Force Material Around Us at Elementary School

Computational thinking is a sophisticated problem-solving approach based on computer science. Implementing computational thinking in elementary schools remains a challenge for education, particularly for teachers and students. Teachers must construct learning experiences using computational thinking to make the learning process more engaging, while students must solve issues logically, systematically, and effectively. This study aims to present an overview of the application of computational thinking in natural and social sciences (IPAS) learning and identify its application in fifth-grade elementary school students. The research employed a qualitative research design with a single one-shot case study. The research subjects were 40 fifth-grade students and four teachers. This investigation was conducted during a single meeting that followed the lesson plan for the force material around us. The findings revealed that (a) analysis of student activity data yielded a percentage of 81.25%; (b) the data analysis of student responses to learning yielded favorable results, approaching 100%. According to interviews and observations of CT (Computational Thinking) in SSI (Socio-scientific Issue) in IPAS subjects, implementing CT learning on the forces around us might bring up aspects of the CT foundation, such as pattern recognition decomposition, abstraction, and algorithms. The learning scenario is that students are requested to assess many types of activities that occur in everyday life. Additionally, students will outline various actions that use force.

Flight Pattern Recognition and Precision Tracking Algorithm for UAV Cluster Targets Imitating Geese Flocks

Flight Pattern Recognition and Precision Tracking Algorithm for UAV Cluster Targets Imitating Geese Flocks

Multi-source partial discharge pattern recognition algorithm based on DCGAN-Yolov5

This study aims to overcome the complication that deep learning-based pattern recognition in partial discharge (PD) diagnosis of gas-insulated switchgear (GIS) can only identify single-source partial discharge but not multi-source partial discharge. Specifically, a GIS multi-source partial discharge detection algorithm based on Deep Convolution Generative Adversarial Networks and You Only Look Once (DCGAN-YOLOv5) is proposed. First, the Phase Resolved Partial Discharge (PRPD) features of multi-source PD are analyzed, a GIS experiment platform is established, and four typical PD defects are simulated. Besides, sample data are collected, and the DCGAN network is used for sample expansion. Then, a YOLOv5 network model is designed, and a spatial and channel attention mechanism is added to the feature extraction network with a positive sample equilibrium strategy. Finally, the effectiveness of the proposed algorithm is verified using laboratory data and field data collected from a 220 kV substation. The experimental results demonstrate that the proposed algorithm can effectively detect the multi-source PD features in PRPD patterns under complex noise and thus successfully identify the types of multi-source PD. The mean Average Precision (mAP) can reach 98.4%. The precision of single-source PD and multi-source PD can reach 95.2% and 89.3%, when testing with field data.

Tracking Emotions Using an Evolutionary Model of Mental State Transitions: Introducing a New Paradigm

Owing to rapid advancements in artificial intelligence, the role of emotion recognition has become paramount in human–computer interaction. Traditional approaches often reduce this intricate task to a mere classification problem by relying heavily on perceptual pattern-recognition techniques. However, this simplification overlooks the dynamic and multifaceted nature of human emotions. According to theories in emotion psychology, existing pattern recognition methods primarily capture external emotional expressions—termed “external emotional energy” (EEE)—rather than the nuanced underlying emotions. To address this gap, we introduce the evolutionary mental state transition model (EMSTM). In the initial phase, EMSTM employs standard pattern-recognition algorithms to extract EEE from multi-modal human expressions. Subsequently, it leverages a mental state transition network to model the dynamic transitions between emotional states, thereby predicting real-time emotions with higher fidelity. We validated the efficacy of EMSTM through experiments on 2 multi-label emotion datasets: CMU Multimodal Opinion Sentiment and Emotion Intensity (CMU-MOSEI) and Ren Chinese Emotion Corpus (Ren-CECps). The results indicate a marked improvement over conventional methods. By synergistically combining principles from psychology with computational techniques, EMSTM offers a holistic and accurate framework for real-time emotion tracking, aligning closely with the dynamic mental processes that govern human emotions.