Abnormal Fluctuations Research Articles

Analyzing Mechanisms of Business Fluctuations involving Time-Varying Structure in Japan: Methodological Proposition and Empirical Study

Abstract In this article, we present a comprehensive analysis of business cycles in Japan from 1975 to 2023, focusing on the mechanisms that underlie economic fluctuations. We develop a novel methodology to construct a comprehensive index of business cycles (CIBC) by integrating 30 indicators commonly used in business cycle analysis in Japan. Unlike traditional approaches, our method simultaneously incorporates all business cycle indicators with adjusted lags, which enhances its precision and scope. We decompose each indicator’s time series using an extended moving linear model, which ensures robustness amid abnormal fluctuations. We propose a two-lag time-varying loading factor model to analyze the lead-lag relationship between the CIBC and each variable using a distribution-free dynamic linear modeling method. In our analysis, we elucidate the complex interactions between key indicators and a common factor, which explains the dynamics of business fluctuations. We conducted a comprehensive empirical study using Japanese business cycle indicator data to demonstrate the performance of the proposed methodology. Despite its limitations, our study provides valuable insights that enable us to understand the mechanisms of business cycles in Japan from the perspective of empirical analysis. Our study lays the groundwork for future research in this area.

Smart Band Enabled AI-System for Mental Health Tracking

The paper presents the development of an AI- enabled smart band designed to continuously monitor mental wellness through the real-time analysis of three key physiological indicators: heart rate, body temperature, and movement patterns. These signals are captured by onboard sensors and transmitted wirelessly to a cloud-based platform using the ESP32/Node MCU module. The system applies AI algorithms to detect variations and correlations in the data that may indicate mental stress, fatigue, or anxiety. Elevated heart rate, abnormal temperature fluctuations, and irregular movement patterns—detected through the accelerometer—serve as critical markers for early- stage mental health concerns. The processed data is visualized on a user-friendly web and mobile dashboard, where users can monitor their mental state trends and receive real-time alerts. When stress thresholds are exceeded, the system delivers instant notifications via app alerts, SMS, or email, encouraging timely interventions. Personalized suggestions, such as light physical activities, deep breathing exercises, or short breaks, are generated based on individual user behavior. The smart band not only provides continuous, non-invasive mental health monitoring but also learns from user data over time using machine learning to improve prediction accuracy. This approach transforms traditional mental health assessment methods by introducing a proactive, AI-driven, and data- centric system for managing everyday stress and emotional well-being.

Numerical simulation of abnormal response characteristics of hydraulic fracturing pumping pressure in strongly heterogeneous volcanic reservoirs

Volcanic reservoirs exhibit significant heterogeneity characteristics influenced by volcanic activity and tectonic evolution. The unclear mechanisms underlying abnormal pressure fluctuations during the pre-pad fluid injection stage in fracturing operations severely constrain reservoir stimulation effectiveness. This study establishes a hydraulic fracture propagation model for heterogeneous reservoirs using the three-dimensional discrete element method, analyzing fracture propagation behaviors and pressure response characteristics to identify dominant factors controlling abnormal pressure responses. Results indicate that bedding, high-angle natural fracture, and localized heterogeneous region with high mechanical strength all impede hydraulic fracture propagation, resulting in pressure surges with maximum increases of 20%, 25%, and 158%, respectively. Due to weak-plane normal stress constraints and additional tortuosity friction, pumping pressure exhibits a continuous rise without decline during hydraulic fracture penetration of weak plane. In contrast, variation in fracture propagation difficulty leads to a rise followed by an abrupt drop in pumping pressure when fracture penetrates localized heterogeneous region. Pumping pressure shows strong sensitivity to Young’s modulus and fracture toughness of localized heterogeneous region. Consequently, the interaction between the hydraulic fracture and localized heterogeneous region with high mechanical strength is the dominant controlling factor leading to abnormal pressure responses during the pre-pad fluid injection stage in volcanic reservoirs.

Ultra-Short-Term Load Forecasting for Extreme Scenarios Based on DBSCAN-RSBO-BiGRU-KNN-Attention with Fine-Tuning Strategy

Extreme scenarios involving abnormal load fluctuations pose serious challenges to the safe and stable operation of power systems. To address these challenges, an ultra-short-term load forecasting model is proposed, specifically designed for extreme conditions. The model combines density-based spatial clustering of applications with noise (DBSCAN), random search Bayesian optimization (RSBO), bidirectional gated recurrent units (BiGRUs), k-nearest neighbor (KNN), and an attention mechanism, enhanced by a fine-tuning strategy to improve forecasting accuracy. Firstly, the original load data are reconstructed weekly, and extreme scenarios are identified using the DBSCAN. Secondly, the RSBO is employed to optimize model parameters within the high-dimensional search space. To further refine performance, the final fully connected layer is fine-tuned to adapt to extreme conditions. Finally, case studies demonstrate that the proposed approach reduces the root mean square error (RMSE) by 12.37% and the mean absolute error (MAE) by 6.73% compared to benchmark models, achieving superior accuracy under all tested extreme scenarios.

Parametric Subnational Population Volatility: The Importance of Abnormal Fluctuations

Parametric Subnational Population Volatility: The Importance of Abnormal Fluctuations

Anomalous pattern recognition based on image recognition in food safety detection

In food processing, foreign matter inevitably contaminates packaged food. To ensure food safety, ray-based detection is used; however, the original images suffer from aberrations and noise that degrade quality and hinder further processing. Thus, images are preprocessed to enhance quality by highlighting key features and suppressing irrelevant ones before abnormal pattern recognition. Following image segmentation, a BP neural network algorithm is applied for foreign object detection. In tests with contaminants such as metal wires, stones, and glass, the algorithm identified distinct abnormal fluctuations at gray levels of 132, 108, and 34, respectively, allowing it to reliably detect foreign objects. Although the practical detection rate reached 100%, occasional misjudgments suggest that further optimization is needed. Overall, this method introduces a novel approach to detecting foreign objects in food and offers promising new strategies for improving food safety monitoring.

Monitoring and Optimization of CFB Bed Temperature in the Flexible Process: A Hybrid Framework of Deep Learning Model and Mechanism Model.

In response to the issues of abnormal bed temperature fluctuations and inefficient combustion that occur during the flexible operation of the circulating fluidized bed (CFB) boiler combustion system, which is characterized by strong coupling, nonlinearity, and large inertia, this paper presents a novel monitoring and optimization approach that integrates a deep learning model with a mechanism model. Specifically, the Informer algorithm is utilized to construct a bed temperature range prediction model, thus enabling the real-time monitoring of the intricate change trends of the in-furnace bed temperature. Considering the poor reliability of combustion optimization target values derived from data model predictions and the constraints on the optimization upper limit, this article further combines the mechanism model to comprehensively determine the combustion optimization target values from the design perspective. The results indicated that the Informer prediction model exhibited a root-mean-square error (RMSE) of 3.385 °C, a mean absolute error (MAE) of 2.45 °C, and a mean absolute percentage error (MAPE) of 0.268% on the rolling test data set, demonstrating high overall prediction accuracy. The steady-state operational data of the generating unit at 300 and 200 MW were selected to validate the mechanism model, with particular emphasis on comparing the distribution trends of pressure and temperature along the height of the furnace. The outcomes revealed good consistency, indicating the model's high accuracy in performance simulation. By integrating the data model with the mechanism model, the target values for optimizing bed temperature performance under steady-state conditions were determined. In comparison to using the data model alone, during the 240 and 280 MW operating conditions, the average thermal efficiency of the boiler increased by 0.19 and 0.13%, respectively. Concurrently, the coal consumption rate for power generation decreased by 0.6707 and 0.4453 g/(kW·h), respectively, and the carbon emissions reduced per kilowatt-hour of electricity generated were 1.6738 and 1.1113 g, respectively.

Instantaneous Frequency Analysis Based on High-Order Multisynchrosqueezing Transform on Motor Current and Application to RV Gearbox Fault Diagnosis

Motor current analysis is useful for ensuring the safety and reliability of electromechanical systems. However, for gearboxes, the commonly used methods of detecting faulty frequency sidebands are easily disturbed by installation errors, inherent harmonics, and fundamental frequency with high amplitude. Aiming at this problem, this study presents instantaneous frequency polarview (IFpolarview), which diagnoses faults based on motor angle and motor current frequency modulation (FM) features. Firstly, to address the problem of the limited analysis order of higher-order synchrosqueezing transform (HSST), the higher-order multisynchrosqueezing transform (HMSST) is introduced to improve the instantaneous frequency (IF) estimation accuracy and reveal the transient fault features from the motor current without further increasing the order and algorithm difficulty. Then, based on the motor angle and accurate motor current IF extracted from HMSST, the IFpolarview is proposed to visualize gear faults through detecting the FM of motor current synchronized with the faulty gear mesh. In the simulation, the IF estimation error of HMSST is 2.51%, which is smaller than other methods. The experimental results show that the HMSST has the smallest Rényi entropy value of 9.13, implying that the most aggregated time–frequency representation (TFR) of the energy is obtained. HMSST can enhance the resolution of fault characteristics, and IFpolarview concentrates the abnormal IF fluctuations with periodicity into a small angular interval, which highlights the fault features and demonstrates greater intuitiveness and reliability in comparison to the frequency sideband detection method.

Fuzzy Particle Filtering Based Approach for Battery RUL Prediction With Uncertainty Reduction Strategies

ABSTRACTThis paper proposes a two‐stage framework that combines uncertainty reduction and predictive modelling to enhance the accuracy of battery Remaining Useful Life (RUL) prediction. In the first stage, a simplified fuzzy optimization learning model is introduced to mitigate uncertainty caused by abnormal capacity fluctuations in battery data. The proposed fuzzy model reconstructs degradation data into a consistent downward trend based on mid‐ and short‐term tendencies of the battery, alleviating abnormal variability and improving suitability for predictive modelling. In the second stage, uncertainty arising during the recursive prediction process of a standalone Transformer model was mitigated through the integration of a particle filter. This approach dynamically manages prediction errors using particles, effectively controlling cumulative errors and enhancing the stability and reliability of long‐term predictions. This methodology can lead to extended battery life and increased operational reliability through accurate RUL prediction. The proposed methodology is validated through experiments using NASA and CALCE battery datasets, demonstrating superior prediction accuracy and stability compared to conventional approaches by systematically reducing uncertainties.

NAD+ Promotes Superovulation of Huaxi Cattle Through Regulation of Cumulus Cell Proliferation and Oocyte Maturation.

Superovulation and embryo transfer are key technologies to improve the reproductive ability of female animals and enhance the efficiency of livestock production. However, poor-quality oocytes or abnormal fluctuations of hormone levels caused by superovulation affect the embryonic development environment, which may lead to a significant decline in the number and quality of transferable embryos, thus reducing the efficiency of superovulation. In this study, nicotinamide adenine dinucleotide (NAD+) was injected into Huaxi cows during the superovulation period to observe the proliferation and apoptosis of transplanted embryos. We examined the proliferation, apoptosis, reactive oxygen species (ROS) and mitochondrial membrane potential of cumulus cells and oocytes directly treated with NAD+ and investigated the potential mechanism of NAD+ to improve the superovulation efficiency by serum metabolomics and single-cell RNA sequencing. The results show that the addition of NAD+ significantly increased the quantity and quality of transferable embryos after superovulation. Differential metabolites during estrus synchronization and embryo flushing are enriched in glycerophospholipid metabolic pathways, suggesting that NAD+ can regulate lipid metabolic pathways. We found that NAD+ optimized the secretion levels of the steroid hormone estradiol (E2) and progesterone (P4) during superovulation by regulating the activity of cumulus cells. In oocytes, we found that NAD+ can inhibit apoptosis, scavenge ROS, and enhance mitochondrial function, thereby promoting oocyte maturation and enhancing embryo developmental potential. In conclusion, NAD+ significantly improved the superovulation ability of Huaxi cattle and provides an effective way for animal husbandry to improve the yield of high-quality embryos.

Association between oxidative balance score and thyroid function and all-cause mortality in euthyroid adults

Abnormal fluctuations in thyroid function within the reference range were strongly associated with increased all-cause mortality. This study aimed to analyze the association between oxidative balance score (OBS) and free thyroxine (FT4) and thyrotropin (TSH) in euthyroid adults, as well as their interrelationships with mortality. 5727 euthyroid adults were selected from the National Health and Nutrition Examination Survey (NHANES). Weighted linear regression investigated the potential association of OBS with FT4 and TSH. In addition, COX proportional hazard models and restricted cubic spline (RCS) were used to investigate the association between OBS, FT4, TSH, and all-cause mortality. The results showed that OBS was negatively associated with serum FT4 concentrations in euthyroid adults (− 2.95%, 95% CI − 5.16%, − 0.92%). Additionally, the all-cause mortality rate was significantly higher in the fourth quartile (Q4) of FT4 compared to the first quartile (Q1) (HR 1.40, 95% CI 1.07–1.85). In the fourth quartile of OBS, the all-cause mortality rate was 31% lower than in Q1 (HR 0.69, 95% CI 0.52–0.92). Mediation analyses indicated that FT4 partially mediated the relationship between OBS and all-cause mortality. These results suggest a significant negative association between OBS and serum FT4, while both OBS and FT4 are strongly associated with mortality. However, the effect of OBS on serum FT4 is relatively limited, and therefore its clinical significance needs to be interpreted objectively.

Analysis of the Influence of Process Parameters on Transverse Flux Induction Heating of Endless-Rolling Strip

This study focuses on the effect of an induction heating device on the entry of a thin strip continuous casting and rolling line. A finite element model for the electromagnetic–thermal coupling of transverse magnetic flux induction heating was developed by adopting COMSOL software 6.1 to systematically investigate the effects of process parameters on the magnetic field, eddy current field, and the transverse temperature distribution of the strip. The results show that when the gap is between 20 mm and 40 mm, the maximum value of magnetic induction in the overheating region at the edges of the strip increases from 0.28 T to 0.35 T and 0.38. When the strip width is 1000 mm, there is an approximately 29% increase in magnetic induction in comparison to a strip with a width of 800 mm, and both eddy current density and temperature exhibit abnormal fluctuations. The maximum temperature difference in the temperature uniformity region at the center of the strip is only 3 °C at different frequencies, and the temperature-rise curves almost completely overlap. With increasing current, the temperature difference between the weak temperature region and the temperature uniformity region at the center widens, indicating a deterioration in temperature uniformity. Meanwhile, the field conditions are simulated using a simplified model of continuous heating. The results indicate that the maximum temperature deviation in the overheating region at the edges of the strip is 6 °C, while the deviation in the temperature uniformity region is 2 °C. Furthermore, the simulation data reveal an average temperature rise of 1156 °C across the width of the strip, with a deviation of 1.4 °C compared to the measured results, which verifies the validity of the proposed model. The analysis results provide a reference basis for designing transverse magnetic flux induction heating devices and optimizing process parameters.

Prediction model and real-time diagnostics of hydraulic fracturing pressure for highly deviated wells in deep oil and gas reservoirs

It is a common occurrence in the fracture processes of deep carbonate reservoirs that the fracturing construction pressure during hydraulic fracturing operation exceeds 80 MPa. The maximum pumping pressure is determined by the rated pressure of the pumping pipe equipment and the reservoir characteristics, which confine the fracture to the target area. When the pump pressure exceeds the safety limit, hydraulic fracturing has to reduce the construction displacement to prevent potential accidents caused by overpressure. Therefore, real-time prediction of the fracturing construction pressure and diagnosis of abnormal fluctuations during hydraulic fracturing of highly deviated wells are indispensable. Based on the well trajectory, pumping process, and string structure of highly deviated wells, a movement interface model for the fracturing fluid at different stages within the wellbore has been established, using the method of computational fluid dynamics. This model analyzes the fluid movement behavior with diverse properties at various fracturing times and determines the relationship between the pressure changes at the leading and trailing edges of fluid movement in each section of the wellbore over time by combining different string structures and preset pumping procedures. The frictional pressure within the wellbore fluid, the hydrostatic fluid pressure, and the near-well friction drags have been calculated and predicted. A real-time prediction model for diagnosing pumping fracturing has been constructed to further comprehend “abnormal” fracturing construction pressures in highly deviated wells. This offers a theoretical foundation for the correct diagnosis and decision-making regarding hydraulic fracturing in highly deviated wells while guiding its smooth implementation in real time.

The application study of deep learning technology in intelligent power supervision systems

Abstract This study explores the application of a combined model incorporating bidirectional long short-term memory networks and gated recurrent units within intelligent power supervision systems, demonstrating its efficacy in recognizing anomalies within power grids. Based on historical operational data of two critical indicators—grid sectional security margin and generation-consumption balance margin—this paper constructs an anomaly detection model. This model effectively processes time series data, promptly identifying equipment malfunctions and abnormal load fluctuations, thereby enhancing the safety and stability of power grids.

Distributed Data Privacy Protection via Collaborative Anomaly Detection

Current anomaly detection methods for charging stations primarily rely on centralized network architectures with federated learning frameworks. However, the rapid increase in the number of charging stations and the expanding scale of these networks impose significant communication traffic loads. Consequently, it is essential to explore the relationship between data aggregation among charging stations and the anomaly detection accuracy at individual stations. In this paper, we address efficient anomaly detection in charging stations and propose a distributed anomaly detection algorithm powered by federated learning. To be specific, we introduce a distributed privacy-preserving data aggregation scheme, where a Transformer model is adopted to effectively smooth abnormal data fluctuations, minimizing disruptions to network aggregation nodes. Furthermore, we develop a distributed federated learning framework incorporating an efficient parameter update method without requiring prior knowledge or a central node. Compared with some existing detection solutions, the proposed approach significantly reduces communication bandwidth requirements while maintaining anomaly detection accuracy and mitigating data isolation issues. Extensive experiments demonstrate that the proposed algorithm not only achieves high accuracy in detecting anomalies in electric vehicle charging stations but also ensures robust user data privacy protection.

Algorithm optimization based on intelligent management of computer electrical equipment: A comprehensive method for PT power monitoring and remote fault indicator

In response to the accuracy limitations and response delays in current power and fault detection of electromechanical equipment, long short-term memory (LSTM) network algorithm was applied for intelligent optimization and management. Firstly, voltage and power data were collected through a potential transformer (PT), and wavelet transform (WT) was applied to remove noise. Fast Fourier transform (FFT) was utilized to extract key features. Secondly, a multi-layer long short-term memory network was designed, and back propagation algorithms and time series power data were used for LSTM network training to analyze abnormal fluctuations and trends in power data in real-time, and adjust threshold settings. Then, combining the model output and historical fault data, a fault mode knowledge base was established. Potential faults were determined through pattern matching, and signals were sent out through remote indicators. Finally, the algorithm model was evaluated. The research results showed that the weighted response time of voltage drop faults was shortened by 3.3%, and the information entropy values of nine experiments were distributed from 5.11 to 5.28. The diagnostic accuracy for frequency drift was improved by 11.1%. The comprehensive algorithm model used can effectively improve the accuracy of power monitoring and response speed. This study optimizes power monitoring and fault diagnosis by applying the LSTM network algorithm, addressing the limitations of existing methods in terms of real-time performance and accuracy. It effectively enhances the fault prediction capability and response speed of power systems, offering significant application value in the fields of smart grids and equipment management.

Comparison of evaluation methods for the dynamic characteristics of railway hydraulic dampers

The EN and Area methods are commonly reported procedures for evaluating dynamic damper characteristics, both predominantly relying on time-domain data of damper force and displacement, with the EN method established as a standard in EN13802. However, these time-domain approaches do not fully utilize all experimental data and have inherent limitations that can impact the accuracy of the results. To address this challenge, this study introduces a calculation method based on the Fast Fourier Transform (FFT), which shifts the analysis from the time domain to the frequency domain, allowing for more reliable determination of phase angle, dynamic stiffness, and damping. Comparative analysis of dynamic test results on typical hydraulic dampers reveals that traditional time-domain methods lead to abnormal fluctuations in dynamic stiffness at low frequencies due to inaccuracies in phase angle calculations. In contrast, the frequency-domain method delivers more reliable results. Among the tested methods (EN, Area, and FFT), the FFT method aligns most closely with the standard EN method, showing discrepancies of less than 10%, and demonstrates greater robustness under noisy conditions. In comparison, the Area method exhibits higher sensitivity to sampling frequency variations, resulting in less reliable outcomes. This study provides practical insights for enhancing damper performance evaluation and recommends the frequency-domain method as a reliable tool for dynamic characteristic analysis, particularly in applications where noise resilience is required, while the EN method remains suitable for low-noise conditions.

An Examination of the Effects of Monetary Policy on the Housing Market and Variations in Asset Prices

With the deepening development of economic globalization, the world has experienced a series of sharp fluctuations in macro-economy. These fluctuations are manifested in the abnormal fluctuations of asset prices and the long-term downturn of stocks, real estate and other industries. Historically, events such as the bursting of Japan's economic bubble in the early 1990s, which led to the collapse of its long-term credit system and the "Heisei Great Depression," and the subprime mortgage crisis that triggered the global financial panic in 2008, have had a profound impact on global financial markets. In this context, how to effectively prevent and cope with the challenges brought by monetary policy macro-control has become the primary task of current research. By studying the sensitivity of monetary policy and asset price, it can effectively predict the crisis and make timely adjustment of related policies. Based on the perspective of financial economics, this paper adopts a two-way analysis combining theoretical analysis and empirical analysis. Specifically, this paper makes an in-depth discussion and analysis by analyzing the annual financial report in which real estate accounts for a relatively large share of local government GDP, combined with the changes in monetary policy.

Radon exhalation rate prediction and early warning model based on VMD-GRU and Similar day analysis

Radon exhalation rate prediction and early warning model based on VMD-GRU and Similar day analysis

The Exploration and Establishment of an Economic Operation Monitoring and Early Warning Model Driven by Big Data

The tobacco industry is a crucial component of the national economy. Enhancing the foresight, relevance, and effectiveness of macroeconomic regulation to ensure that the cigarette economy operates within a reasonable range is vital for improving macroeconomic alignment and stabilizing its contribution. This study proposes a monitoring index system for the cigarette economy. The SARIMA model is applied to forecast key indicators scientifically, while algorithms like the BP neural network and the prosperity index establish a big data-driven early warning system for cigarette economic operations. This system identifies real-time trends, detects abnormal fluctuations, shifts from reactive regulation to proactive prevention, and provides data-driven support for management decisions.