Early Warning Research Articles

Monitoring and analysis of physical activity and health conditions based on smart wearable devices

The rapid growth of the Internet of Things (IoT) brings sweeping changes in various industries. Healthcare industries have become a prime example where the Internet of Healthcare Things (IoHT) is making significant progress, particularly in how we approach real-time patient care. Traditional systems for monitoring older people and people with special needs are frequently expensive, require a large workforce, and fall short of providing real-time data. This paper introduces the “3-Tier Health Care Architecture,” an integrated approach to mitigating these issues. This architecture capitalizes on IoHT technologies and is constructed around three principal tiers: Sensor, Fog, and Cloud. The Sensor Tier employs Health Metrics Acquisition Units (HMAUs) fitted with an nRF5340 Development Kit, capturing an extensive range of health-related metrics via wearable sensors. These metrics are then relayed to the Local Processing Units (LPUs) in Fog Tier, which operates on Raspberry Pi Zero 2 W microprocessors for the initial data processing before forwarding to the cloud. The Cloud Tier uses a hybrid CNN-LSTM Machine Learning (ML) model to perform Real-Time Healthcare Monitoring (RTHM) status assessments and includes an Early Warning System for immediate alert issuance. The proposed architecture is resilient, scalable, and efficient, serving as a fortified and all-encompassing solution for RTHM. This enables quick medical interventions, thus elevating healthcare quality and potentially life-saving.

A convolutional neural network based on an evolutionary algorithm and its application

PM2.5 concentration predictions can provide air pollution control, management, and early warning. However, the PM2.5 data with high-dimensionality, complexity, and dynamics pose a great challenge to achieve optimal prediction results. Convolutional Neural Networks (CNNs) has unique advantages in processing complex data and contributes to state-of-the-art performances. However, designing the architecture and selecting the learning rate for CNN are time-consuming and requires prior knowledge. Evolutionary algorithms, with the advantages of global convergence, ergodicity, robustness and adaptability, are the most commonly used methods to design the optimal framework for CNNs. Therefore, to improve the predictive performance of CNNs, this paper proposes an improved CNN method (EBRO-ICNN) which employs the enhanced battle royale optimization (EBRO) algorithm and proportional-derivative (PD) control. Firstly, the EBRO algorithm with multistrategy collaborative optimization is introduced, and validated by CEC2017 benchmark functions, which demonstrates EBRO strong global exploration capability, fast convergence speed, and low time complexity. Next, PD control is applied to adjust the learning rate of CNN (ICNN) dynamically, which improves the efficiency and stability of the network training process. At the same time, the ICNN model is optimized using the EBRO algorithm, which can reduce the human interference, generate the optimal framework automatically and enhance the prediction accuracy effectively. Finally, a private air quality dataset and two public datasets are utilized to evaluate the performance of the EBRO-ICNN model, considering 3 error evaluation metrics and 7 prediction comparison models. The experimental results demonstrate that the EBRO-ICNN model exhibits great accuracy and stability.

The rapid detection method of brown tide algae concentration based on unthresholded recurrence plots and improved broad learning system

In recent years, there have been frequent outbreaks of harmful algal blooms (HAB) in coastal areas, which have caused serious economic losses to the local community. Therefore, accurate and rapid prediction of microalgal concentrations is necessary for early warning and countermeasures before the occurrence of HAB. This paper presents a model for predicting microalgae concentration based on unthresholded recurrence plots (UTRPs) combined with an improved broad learning system (BLS). Spectral data acquisition of algae species at different concentrations using l-induced fluorescence spectroscopy. Then, the 1D spectral data are dimensionally lifted by UTRPs transformation, recurrence plots (RPs) transformation can fully extract the internal information of 1D sequence data, and at the same time, UTRPs avoids the influence of artificially selected thresholds on the feature transformation results of traditional (RPs). Finally, a lightweight flat network BLS was used for microalgae concentration regression prediction, at the same time the BLS regularization method was improved. When comparing the two most commonly used deep learning regression models and integrated learning models, UTRPs and elastic net (UTRPs-ENBLS) achieves convincing results.

Gut Bacteria Provide Genetic and Molecular Reporter Systems to Identify Specific Diseases.

With genetic information gained from next-generation sequencing (NGS) and genome-wide association studies (GWAS), it is now possible to select for genes that encode reporter molecules that may be used to detect abnormalities such as alcohol-related liver disease (ARLD), cancer, cognitive impairment, multiple sclerosis (MS), diabesity, and ischemic stroke (IS). This, however, requires a thorough understanding of the gut-brain axis (GBA), the effect diets have on the selection of gut microbiota, conditions that influence the expression of microbial genes, and human physiology. Bacterial metabolites such as short-chain fatty acids (SCFAs) play a major role in gut homeostasis, maintain intestinal epithelial cells (IECs), and regulate the immune system, neurological, and endocrine functions. Changes in butyrate levels may serve as an early warning of colon cancer. Other cancer-reporting molecules are colibactin, a genotoxin produced by polyketide synthetase-positive Escherichia coli strains, and spermine oxidase (SMO). Increased butyrate levels are also associated with inflammation and impaired cognition. Dysbiosis may lead to increased production of oxidized low-density lipoproteins (OX-LDLs), known to restrict blood vessels and cause hypertension. Sudden changes in SCFA levels may also serve as a warning of IS. Early signs of ARLD may be detected by an increase in regenerating islet-derived 3 gamma (REG3G), which is associated with changes in the secretion of mucin-2 (Muc2). Pro-inflammatory molecules such as cytokines, interferons, and TNF may serve as early reporters of MS. Other examples of microbial enzymes and metabolites that may be used as reporters in the early detection of life-threatening diseases are reviewed.

Real-time earthquake magnitude estimation via a deep learning network based on waveform and text mixed modal

Rapid and accurate earthquake magnitude estimations are essential for earthquake early warning (EEW) systems. The distance information between the seismometers and the earthquake hypocenter can be important to the magnitude estimation. We designed a deep-learning, multiple-seismometer-based magnitude estimation method using three heterogeneous multimodalities: three-component acceleration seismograms, differential P-arrivals, and differential seismometer locations, with a specific transformer architecture to introduce the implicit distance information. Using a data-augmentation strategy, we trained and selected the model using 5365 and 728 earthquakes. To evaluate the magnitude estimation performance, we use the root mean square error (RMSE), mean absolute error (MAE), and standard deviation error (ϭ) between the catalog and the predicted magnitude using the 2051 earthquakes. The model could achieve RMSE, MAE, and ϭ less than 0.38, 0.29, and 0.38 when the passing time of the earliest P-arrival is 3 s and stabilize to the final values of 0.20, 0.15, and 0.20 after 14 s. The comparison between the proposed model and model ii, which is retrained without the specific architecture, indicates that the architecture contributes to the magnitude estimation. The P-arrivals picking error testing indicates the model could provide robust magnitude estimation on EEW with an absolute error of less than 0.2 s.Graphical

Enhanced feature combinational optimization for multivariate time series based dynamic early warning in power systems

Multivariate time series (MTS) learning-based dynamic early warning (DEW) shows promising potential in preventing power systems from underlying risks. However, as a fundamental and essential step for constructing DEW models, performing efficient feature selection (FS) within high-dimensional candidate features (HDCF) of power grids is challenging, especially considering the MTS format and security-dominated characteristics of electric utilities. This paper proposes an enhanced feature combinational optimization (FCO) method to select beneficial variables for MTS-based DEW models within the HDCF of power grids. First, to identify features that contain valuable temporal patterns for DEW, a wrapper FCO framework combined MTS learning and evolutionary computation is designed. Then, to improve the FCO efficiency and solution quality when facing HDCF, a maximal information coefficient (MIC) based enhancement scheme is developed, semi-guiding the FCO initialization and evolution. Finally, by integrating MIC and iterative search, hybrid end-to-end FS is achieved. Experiments on the synthetic data from the IEEE 39-bus system and the real-world data acquired from a provincial power grid in China show that, via the MIC semi-guidance, the proposed method possesses higher FCO efficiency and better results within HDCF compared to wrapper benchmarks. Compared to filter benchmarks, it is more compatible with supervised MTS and can consider the practical performance of features by interacting with the embedded learning models. Results also verify that, with the features from the proposed method, more accurate DEW models can be obtained to predict the target event in early stages.

Nutrient dynamics in wetland systems associated with hydrological and anthropogenic variations in the south of Samborombón Bay, Argentina

Wetlands provide numerous ecosystem services to the environment, such as nutrient removal and storage. The aim of this work is to evaluate the nutrient dynamics in different sectors within wetland systems in the south of the Samborombón Bay (Argentina) based on hydrological and/or anthropogenic variations. For this purpose, the hydrological features of the wetland were defined through the analysis of satellite images, precipitation and tidal data, and field surveys. Three sectors were identified in the wetland: one with tidal influence, another which is dependent on rainfall, and another that receives inputs from rainfall and from a sewage effluent treatment plant. In order to analyze the nutrient dynamics, samples of surface water, groundwater, and sediments were collected from these sectors. Measurements of pH and electrical conductivity were determined in situ in water samples, while the concentration of inorganic forms of phosphorus and nitrogen, soluble reactive silica, and organic carbon were determined through laboratory analysis. Also, bioavailable phosphorus, organic matter, pH, and electrical conductivity were determined in the sediment samples collected. Statistical analysis of the data reveals differences between the sectors and allows the interpretation of the dynamics of the studied components in the wetland. Electrical conductivity distinguishes the intertidal sectors of the wetland while components associated with P and N discriminate the sectors with inputs from the sewage effluent treatment plant. On the other hand, soluble reactive silica, organic carbon, and organic matter do not seem to be influenced by the tide or effluent discharge. This study demonstrates that the studied wetland works as a nutrient retention area, providing ecosystem services to local inhabitants. Although these services can be utilized, they require a continuous monitoring over time to provide an early warning in case the variations in P and N cycles could lead to eutrophication or wetland degradation.

An integrative warning-protection shear thickened composite sponge towards sensing performance and impact resistance with excellent flame retardant

The mechanical shock and thermal damage generally exist together in hazardous environment, whereas traditional single protective materials are difficult to achieve all-round protection under extreme environment. Herein, to realize the integration of multifunctional abilities about impact resistance, flame-retardant and sensing performance, a “solid-liquid host-guest” structural shear thickened composite sponge is proposed, which flame-retardant shear thickening fluids with an extraordinary thickening ratio of 2183.95 % continuously disperse in 3D sponge matrix. The resulting composite sponge effectively attenuates the impact force by nearly 80 % with the thickness of only 4 mm via a cooperation between the effect of shear hardening and structure densification. Meanwhile, after burning on the alcohol flame for 12 s, the composite sponge with 7 wt.% fire-retardant additive maintains a complete morphology and realizes self-extinguishing with a reliable flame-retardant ability. Attributed to the good electrical conductivity of shear thickening fluid, the composite sponge serves favorably in human body monitoring, impact sensing and fire warning. In a word, this multifunctional lightweight composite sponge is expected to realize the integration of early warning and protection, which is a competitive candidate material for the intelligent protection in complex environments.

Detection and recognition of the invasive species, Hylurgus ligniperda, in traps, based on a cascaded convolution neural network.

Hylurgus ligniperda, an invasive species originating from Eurasia, is now a major forestry quarantine pest worldwide. In recent years, it has caused significant damage in China. While traps have been effective in monitoring and controlling pests, manual inspections are labor-intensive and require expertise in insect classification. To address this, we applied a two-stage cascade convolutional neural network, YOLOX-MobileNetV2 (YOLOX-Mnet), for identifying H. ligniperda and other pests captured in traps. This method streamlines target and non-target insect detection from trap images, offering a more efficient alternative to manual inspections. Two cascade convolutional neural network models were employed in two stages to detect both target and non-target insects from images captured in the same forest. Initially, You Only Look Once X (YOLOX) served as the target detection model, identifying insects and non-insects from the collected images, with non-insect targets subsequently filtered out. In the second stage, MobileNetV2, a classification network, classified the captured insects. This approach effectively reduced false positives from non-insect objects, enabled the inclusion of additional classification terms for multi-class insect classification models, and utilized sample control strategies to enhance classification performance. Application of the cascade convolutional neural network model accurately identified H. ligniperda, and Mean F1-score of all kinds of insects in the trap was 0.98. Compared to traditional insect classification, this method offers great improvement in the identification and early warning of forest pests, as well as provide technical support for the early prevention and control of forest pests. © 2024 Society of Chemical Industry.

HATE SPEECH ON SOCIAL MEDIA: INDONESIAN NETIZENS’ HATE COMMENTS OF PRESIDENTIAL TALK SHOWS ON YOUTUBE

Hate comments on digital platforms, particularly social media, pose a significant threat to online investigations and present a complex issue for linguistic culture. Previous studies have mainly focused on the analysis of moderation strategies regarding this type of language use on the internet. In contrast, little attention has been given to identifying which hate comments on the internet are considered problematic. The current study investigates the phenomena of Indonesian netizens’ hate comments regarding the presidential election in 2024 on YouTube talk shows uploaded in 2023. By analyzing 315 datasets using Nvivo software, it shows that early warning (43%) is mostly employed by Indonesian netizens, followed by dehumanization and demonization (21%), violence and incitement (19%), and offensive language (17%) respectively. Moreover, it is found that anonymity and personality traits are the factors contributing to hate comments related to the 2024 presidential election. The factors influence Indonesian netizens in early-warning hate comments. The implications of the present study highlight the importance of examining the online behavior and language usage in online communication among Indonesian netizens. This understanding can help in mitigating hate comments on online platforms and in society as a whole.

Anomaly Detection Algorithm for Urban Infrastructure Construction Equipment based on Multidimensional Time Series

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.

A Time-Frequency Domain Mixed Attention-Based Approach for Classifying Wood-Boring Insect Feeding Vibration Signals Using a Deep Learning Model.

Wood borers, such as the emerald ash borer and holcocerus insularis staudinger, pose a significant threat to forest ecosystems, causing damage to trees and impacting biodiversity. This paper proposes a neural network for detecting and classifying wood borers based on their feeding vibration signals. We utilize piezoelectric ceramic sensors to collect drilling vibration signals and introduce a novel convolutional neural network (CNN) architecture named Residual Mixed Domain Attention Module Network (RMAMNet).The RMAMNet employs both channel-domain attention and time-domain attention mechanisms to enhance the network's capability to learn meaningful features. The proposed system outperforms established networks, such as ResNet and VGG, achieving a recognition accuracy of 95.34% and an F1 score of 0.95. Our findings demonstrate that RMAMNet significantly improves the accuracy of wood borer classification, indicating its potential for effective pest monitoring and classification tasks. This study provides a new perspective and technical support for the automatic detection, classification, and early warning of wood-boring pests in forestry.

Dual straight-wing FBG accelerometer for low-frequency vibration measurement

Fiber optic accelerometers have a wide range of applications in low-frequency vibration measurement, such as in earthquake monitoring, disaster early warning, underground resource exploration, and anti-seismic monitoring projects for critical structures like dams and bridges. In response to the current issues of Fiber Bragg Gratings (FBG) accelerometers being insensitive to low-frequency vibration signals and challenging to apply in engineering, a dual straight-wing FBG accelerometer for low-frequency vibration measurement is proposed. Firstly, a sensor model is established and theoretically analyzed. Secondly, the impact of the sensor's main structural parameters on sensitivity and natural frequency is analyzed, and the sensor is simulated using ANSYS. Finally, a prototype of the sensor is developed, and a low-frequency vibration testing system is set up to test the sensor's performance. Experimental results show that the accelerometer has a natural frequency of 39.1 Hz, operates in the 5–22 Hz frequency band, and has a sensitivity of 171.7 p.m./g, a dynamic range of 69.82 dB, and a transverse interference resistance of less than 9.5%. This research provides a new reference for the development of similar sensors.

Rockburst early warning based on microseism and critical point theory

During the tunnel construction, rockbursts occurred, posing a risk to safety. To mitigate this risk, a microseismic monitoring system was installed in the tunnel to provide early warning of rockbursts. After conducting research on the evolution of microseismic events in terms of number, moment magnitude, and energy within a specific time frame, an approach for rockbursts risk assessment based on microseismic moment magnitude was proposed. The critical point theory was utilized to quantitatively analyze the energy release process of the surrounding rock before intense rockbursts. This analysis revealed that the energy release of the surrounding rock before rockbursts can be divided into three stages: stable energy release stage, transitional energy release stage and accelerated energy release stage. The obtained process, derived from the critical point theory, was compared and analyzed alongside the evolution process of value of b, cumulative apparent volume, and energy index.

Anomaly Detection for Landslide Displacement Monitoring Data Based on TCN-Transformer

In view of the high false alarm rate of landslide early warning, this paper proposes a classification method of landslide displacement monitoring data based on TCN and attention mechanism to identify normal data and abnormal data. First, the original monitoring data is processed and denoised to improve the data quality. Then, TCN is used to capture the long-term dependence of time series data and highlight key features in combination with attention mechanism. Finally, the landslide displacement monitoring data can be accurately classified by building a classification model to find out the abnormal data and reduce the false alarm rate of early warning. Experiments show that this method can recognize abnormal displacement monitoring data well while ensuring a high recall rate, and reduce the false alarm rate of early warning.

Systemic Financial Risk Forecasting with Decomposition–Clustering-Ensemble Learning Approach: Evidence from China

Establishing a scientifically effective systemic financial risk early warning model is of great significance for prudently mitigating systemic financial risks and enhancing the efficiency of financial supervision. Based on the measurement of systemic financial risk and the network sentiment index of 47 financial institutions, this study adopted the “decomposition–reconstruction–integration” approach, utilizing techniques such as extreme-point symmetric empirical mode decomposition (ESMD), empirical mode decomposition (EMD), variational mode decomposition (VMD), hierarchical clustering, fast independent component analysis (FastICA), attention mechanism, bidirectional long short-term memory neural network (BiLSTM), support vector regression (SVR), and their combination, to construct a systemic financial risk prediction model. The empirical results demonstrate that decomposing and reconstructing relevant indicators before predicting systemic financial risks can enhance prediction accuracy. Among the proposed models, the ESMD-HFastICA-BiLSTM-Attention model exhibits superior performance in systemic financial risk early warning.

Characteristics Analysis of Influence of Multiple Parameters of Mixed Sea Waves on Delay–Doppler Map in Global Navigation Satellite System Reflectometry

Feature capture and recognition of sea wave components in radar systems especially in global navigation satellite system reflectometry (GNSS-R) using signal processing approaches or computer simulative methods has become a research hotspot in recent years. At the same time, parameter inversion of marine phenomena from the discovered characteristics plays a significant role in monitoring and forewarning the different components of sea waves. This paper aims to investigate the impact of multiple parameters, such as the wind speed, directionality variable, wave amplitude, wave length, and directions of sea wave components, on the delay waveform of the delay–Doppler map (DDM). Two types of wind waves and the 2-D sinusoidal sea surface are chosen to be analyzed. By comparing and analyzing the discrepancy of delay waveforms under different conditions, it can be concluded that the increased MSS which arises from the increase in the roughness of the sea surface can lead to the difference in the peak value or trial edges exhibited in delay waveforms. The values of delay waveforms at zero chip along the increasing direction of long-crest wind waves exhibit the periodic spikes shape, which is the opposite of the short-crest wind waves, and the fluctuation of the periodic profiles decreases with the increase in the amplitude of waves. The results and conclusions can provide a foundation for the parameter inversion, tracking, and early warning of anomalous formations of waves in bistatic radar configuration.

Characteristics of drug resistance and biofilm formation in carbapenem-resistant Acinetobacter baumannii in hospitalized children

To study the distribution, drug resistance, and biofilm characteristics of carbapenem-resistant Acinetobacter baumannii (CRAB) isolated from hospitalized children, providing a reference for the prevention and treatment of CRAB infections in hospitalized children. Forty-eight CRAB strains isolated from January 2019 to December 2022 were classified into epidemic and sporadic strains using repetitive extragenic palindromic sequence-based polymerase chain reaction. The drug resistance, biofilm phenotypes, and gene carriage of these two types of strains were compared. Both the 22 epidemic strains and the 26 sporadic strains were producers of Class D carbapenemases or extended-spectrum β-lactamases with downregulated outer membrane porins, harboring the VIM, OXA-23, and OXA-51 genes. The biofilm formation capability of the sporadic strains was stronger than that of the epidemic strains (P<0.05). Genes related to biofilm formation, including Bap, bfs, OmpA, CsuE, and intI1, were detected in both epidemic and sporadic strains, with a higher detection rate of the intI1 gene in epidemic strains (P<0.05). CRAB strains are colonized in the hospital, with sporadic strains having a stronger ability to form biofilms, suggesting the potential for forming new clonal transmissions in the hospital. Continuous monitoring of the epidemic trends of CRAB and early warning of the distribution of epidemic strains are necessary to reduce the risk of CRAB infections in hospitalized children.

Lactate is Associated with Long-Term Mortality in Hospitalized Patients with COVID-19: A Retrospective Cohort Study

Introduction: A recent pandemic caused by Coronavirus 2019 (COVID-19) caused mild and severe systemic organ involvement that led to the death of enormous numbers of people. The prevalence of the disease has declined over the past years, but concerns about upcoming mutations remain. In this study, the relationship between lactate levels and mortality at different times was investigated. Methods: In the present retrospective study, 228 hospitalized patients with COVID-19 were included according to the inclusion and exclusion criteria. A modified National Early Warning Score 2 (NEWS2) was used to determine the severity of the patients' conditions. Follow-up of patients, if discharged alive, has been done from hospitalization until March, 2022. Data were analyzed using SPSS version 22, and p &lt; 0.05 was considered significant. Results: Lactate levels (2.88 ± 2.37 in the dead group versus 1.68 ± 1.33 in the living group) have a significant relationship with mortality at hospitalization (p &lt; 0.001). Furthermore, higher lactate levels during hospitalization (p &lt; 0.001, HR = 2.960, 95%CI =4.255-2.58) and follow-up (p &lt; 0.001, HR = 2.960, 95%CI =4.255 to 2.58) increased the mortality risk ratio by more than two-fold. Conclusion: This study reported that initial lactate levels at admission predict COVID-19 patients' mortality at hospitalization and follow-up. However, further research is needed in this area

Estimating soil water content from thermal images with an artificial neural network

Data of soil water content measured at regular time intervals could be used to set an irrigation schedule for farmland. Changes in soil water content could also provide an early warning of potential slope slippage. An artificial neural network is proposed for detecting soil water content from five variables: soil and water temperature changes after sunlight exposure, plant coverage rates, fine-grained soil content, and effective soil particle size. Temperature changes were detected using thermal images taken using an unmanned aerial vehicle equipped with a thermal camera in the morning and at noon. Plant coverage was estimated using visible light photograph, and the fine-grained soil content, effective particle size, and soil water content were determined from laboratory tests of soil samples. An optimized neural network was established by analyzing the performance of various neural network parameters. The mean square error of the optimized neural network reached a minimum value of 0.10061 in the validation data set at the 12th epoch, the training was stopped to avoid overfitting. The correlation coefficients for the training, validation, testing, and overall data sets were 0.880, 0.772, 0.810, and 0.847 respectively. The target values and the predicted values of the neural network are highly correlated. Thus, the proposed neural network is effective for predicting soil water content. A relative importance analysis revealed that soil and water temperature changes after sunlight exposure had the greatest influence on the neural network’s results.