Series Of Transformations Research Articles

Technological Influence on Communications and Society: All Roads Lead to AI

This issue covers the history of AI, the role of AI in transforming public relations, political communications on social media, the relationship between post-culture and Generation Z, ethical dilemmas of social media use, the dynamics of news value, the narrative transformation of video game-based television series, and the ethical use of AI in marketing.

Positive-definite Converging Kernel Estimation of Long-run Variance

Kernel estimators have been popular for decades in long-run variance estimation. To minimize the loss of efficiency measured by the mean-squared error in important aspects of kernel estimation, we propose a novel class of converging kernel estimators with three major properties: (1) the optimal bandwidth choice is model-free; (2) positive-definiteness is ensured through a principle-driven aggregation technique with no loss of theoretical efficiency; and (3) potentially misspecified prewhitening models and transformations of the time series do not harm the asymptotic efficiency. A shrinkage prewhitening transformation is proposed for more robust finite-sample performance. The estimator has a positive bias that diminishes with the sample size so that it is more conservative compared with the typically negatively biased classical estimators. The proposal improves upon standard kernel functions and can be well generalized to the multivariate case. We discuss its performance through simulation results and a real-data application in the forecast breakdown test.

TCM: An efficient lightweight MLP-based network with affine transformation for long-term time series forecasting

Time series forecasting (TSF) involves extracting underlying patterns from past information to predict future sequences over a specific period. Extending the prediction length of time series and improving the prediction accuracy have always been challenging tasks. Autoregressive prediction methods based on Markov chains tend to accumulate errors over time. Although Transformer-based methods with various self-attention mechanisms have shown some improvements, they require higher memory and computational resources. In this work, we present an effective MLP-based TSF framework named TCM, which models the sequence and channel dependencies separately using Token MLP and Channel MLP. Additionally, we employ the Affine Transformation to replace layer normalization or batch normalization, leading to substantial enhancements in both accuracy and inference speed. Compared to current state-of-the-art long-term time series forecasting models, TCM achieves 6.0% relative improvement on seven real-world datasets, including electricity, weather, and illness domains. The TCM model, characterized by its efficiency and lightweight architecture, also makes it suitable for scenarios with high real-time requirements.

A two-step approach for damage identification in bridge structure using convolutional Long Short-Term Memory with augmented time-series data

This paper presents a novel two-step approach to identifying structural damages in bridge structure through the integration of 1D Convolutional Neural Network (1DCNN) and Long Short-Term Memory (LSTM) networks, enhanced by the augmentation and transformation techniques using Symbolic Aggregate approXimation (SAX) for time-series data analysis. In the first step, the time-series data of the bridge is diversified and quantified by augmentation techniques to make the model more robust and increase its generalization capabilities. After that, SAX is implemented to reduce the volume and categorize time series data through the transformation of continuous time series into discrete symbols, thereby decreasing the size of the data for more efficient training performance. In the second step, an advanced DL model combining 1DCNN and LSTM is proposed to tackle the damage identification problems of the processed data. By leveraging the strengths of CNNs in feature extraction and LSTMs in sequence learning, combined with advanced techniques for data augmentation, our methodology offers a robust solution not only for improving the model's training process but also for enabling it to learn from a more diverse and comprehensive dataset that mimics different damage scenarios, allowing more accurate detection of damages within bridge structures. Validation of the proposed method is conducted using time-series data collected from Chuong Duong Bridge structure. The effectiveness of the proposed method is compared with other models, such as 1DCNN, LSTM, and the combined 1DCNN-LSTM. The results show that the proposed 1DCNN-LSTM-SAX outperforms the other methods in terms of accuracy and, thus, can be used extensively to deal with the damage identification problems of bridges using time-series data.

Analysis of signals from air conditioner compressors with ordinal patterns and machine learning

Most machines are equipped with devices that monitor their operation. Air conditioners, in particular, are routinely monitored through various measurements. A desirable outcome of this monitoring is identifying when the device will likely require maintenance. In this study, we present the use of Ordinal Patterns, a symbolic transformation of time series, which allows for the visual assessment of the type of operation. Ordinal Patterns are chosen because they can transform intricate time series into simple and intuitive symbolic representations. The technique is visually appealing, generating points on a plane whose positions reveal hidden dynamics. This approach makes it easier to identify recurring or abnormal patterns in machine operations that may indicate wear or impending failure. Additionally, Ordinal Patterns allow for precise and understandable visualization of operational data, making interpreting results more accessible for professionals who may not be experts in data analysis. We compare two machines under different operational conditions with six measured variables. We analyze the expressiveness of the Ordinal Patterns and identify those variables that best differentiate the two machines. Furthermore, we incorporate machine learning algorithms, such as Artificial Neural Networks, Support Vector Machines, and Decision Trees, to evaluate and validate the effectiveness of Ordinal Patterns as discriminative features. Integrating machine learning methods with a symbolic transformation offers a robust approach for the early and accurate diagnosis of potential failures, enhancing the predictive maintenance of air conditioning equipment.

Multifractal properties of Covid-19 data

Multifractal detrended moving average analysis (MFDFA) has been performed on time series of three different epidemiological indices, namely daily number of new cases per million population, reproduction rate and positive rate, recorded during the Covid-19 pandemic. We consider India, Japan, USA and UK for this study. For all the series the scaling behavior of MFDMA fluctuation functions has been examined, and the multifractal variables like the generalised Hurst exponent spectra and singularity spectra have been calculated. It is found that the time series of the Covid-19 indices are of multifractal nature. The degree of multifractality of the reproduction rate series is significantly weaker than the other two series for all the countries. To find out the source(s) of multifractality, we generate a set of one hundred shuffled series and Amplitude-Adjusted Fourier Transform surrogate series for each of the original series. Comparing the results obtained from the original series with those from the shuffled and surrogate series we understand that the main source of multifractality in Covid-19 data is long-range temporal correlation among the series values.

Evaluating the Effectiveness of Time Series Transformers for Demand Forecasting in Retail

This study investigates the effectiveness of Transformer-based models for retail demand forecasting. We evaluated vanilla Transformer, Informer, Autoformer, PatchTST, and temporal fusion Transformer (TFT) against traditional baselines like AutoARIMA and AutoETS. Model performance was assessed using mean absolute scaled error (MASE) and weighted quantile loss (WQL). The M5 competition dataset, comprising 30,490 time series from 10 stores, served as the evaluation benchmark. The results demonstrate that Transformer-based models significantly outperform traditional baselines, with Transformer, Informer, and TFT leading the performance metrics. These models achieved MASE improvements of 26% to 29% and WQL reductions of up to 34% compared to the seasonal Naïve method, particularly excelling in short-term forecasts. While Autoformer and PatchTST also surpassed traditional methods, their performance was slightly lower, indicating the potential for further tuning. Additionally, this study highlights a trade-off between model complexity and computational efficiency, with Transformer models, though computationally intensive, offering superior forecasting accuracy compared to the significantly slower traditional models like AutoARIMA. These findings underscore the potential of Transformer-based approaches for enhancing retail demand forecasting, provided the computational demands are managed effectively.

Ultra-Short-Term Wind Power Prediction Based on the ZS-DT-PatchTST Combined Model

When using point-by-point data input with former series models for wind power prediction, the prediction accuracy decreases due to data distribution shifts and the inability to extract local information. To address these issues, this paper proposes an ultra-short-term wind power prediction model based on the Z-score (ZS), Dish-TS (DT), and Patch time series Transformer (PatchTST). Firstly, to reduce the impact of data distribution shift on prediction accuracy, ZS standardization is applied to both training and testing datasets. Additionally, the DT algorithm, which can self-learn the mean and variance, is introduced for window data standardization. Secondly, the PatchTST model is employed to convert point input data into local-level input data. Feature extraction is then performed using the multi-head attention mechanism in the Encoder layer and a feed-forward network composed of one-dimensional convolution to obtain the prediction results. These results are subsequently de-standardized using DT and ZS to restore the original data amplitude. Finally, experimental analysis is conducted, comparing the proposed ZS-DT-PatchTST model with various prediction models. The proposed model achieves the highest prediction accuracy, with a mean absolute error of 5.95 MW, a mean squared error of 10.89 MW, and a coefficient of determination of 97.38%.

A new multi-modal time series transformation method and multi-scale convolutional attention network for railway wagon bearing fault diagnosis

To align with the evolving trends in intelligent railway wagon operation and maintenance and to enhance the precision of railway wagon bearing fault diagnosis, this paper introduces a novel method for bearing fault diagnosis. The method comprises two key innovations. Firstly, a multi-modal time series transformation method is proposed. This method extracts time series data from the original time domain signals via self-adaptive ensemble empirical mode decomposition with adaptive noise, transforms them into 2D matrices, and captures inter- and intra-period information relationships through convolution. Secondly, a multi-scale convolutional attention network is introduced, enriching fault information by utilizing parallel multi-scale convolution for down-sampling. To prevent feature loss, sliding convolution is adopted instead of pooling. Additionally, the model incorporates the convolutional block attention module to focus on critical information. Experimental validation conducted in a laboratory using a self-developed railway wagon bearing dynamic performance tester demonstrates high diagnostic accuracy and strong overall performance. The method’s generalizability is further confirmed through validation using publicly available datasets. This method could find practical use in railway maintenance, improving the accuracy of bearing fault diagnosis, and making operations more efficient.

Benchmarking Approaches: Time Series Versus Feature-Based Machine Learning in ECG Analysis on the PTB-XL Dataset.

Extensive research has been conducted on time series and tabular data in the context of classification tasks, considering their distinct data domains. While feature extraction enables the transformation of series into tabular data, direct comparative comparisons between these data types remain scarce. Especially in the domain of medical data, such as electrocardiograms (ECGs), deep learning faces challenges due to its lack of easy and fast interpretability and explainability. However, these are crucial aspects for a wide and reliable adoption in the field. In our study, we assess the performance of XGBoost and InceptionTime on ECG features and time series data respectively. Our findings reveal that features extracted from ECG signals not only achieve competitive performance but also retain advantages during training and inference. These advantages encompass accuracy, resource efficiency, stability, and a high level of explainability.

Smart grid stability prediction model using two-way attention based hybrid deep learning and MPSO

In smart grid management, precise stability prediction is a complicated task that adds to the effective allocation of resources with grid stability. Specifically, demand-side management is considered an essential element of the overall Smart Grids system. Hence, predicting future energy demands is crucial to regulating consumption by aligning utility offerings with consumer demand. This research presents a hybrid deep learning model (Convolutional Neural Network [CNN] with Bi-LSTM) with a two-way attention method and a multi-objective particle swarm optimization method (MPSO) for short-term load prediction from a smart grid. The proposed hybrid model utilizes a two-way attention method at its encoding and decoding stages, in which an encoding attention layer helps to recognize all the essential features from an input vector, and a decoding attention layer helps to resolve the fixed context vector problem by offering better memory capacity. A CNN and Bi-LSTM are used to capture the essential features from the dataset. We also utilize a t-Nearest Neighbours algorithm to pre-process the initial dataset. An MPSO method combines the features of CNN and Bi-LSTM methods, resulting in better prediction accuracy. As far as we know, it is the first work to suggest a dynamic short-term load prediction model that considers different significant features and enables precise predicting outcomes. The performance of the proposed model and existing well-known deep learning models such as Recurrent Neural Network, Gated Recurrent Unit, Long Short-Term Memory (LSTM), Time Series Transformer, CNN-LSTM and various performance measuring parameters MAE, MSE, MAPE and RMSE are calculated on online UCI dataset (Electrical Grid Stability Simulated Dataset). The proposed hybrid model achieved a better prediction result, which proves the efficiency of the proposed model.

Enhanced urban PM2.5 prediction: Applying quadtree division and time-series transformer with WRF-chem

Accurately predicting air quality levels in urban landscapes is a key environmental and public health challenge due to the complex interactions between pollutant emissions, atmospheric chemistry, and meteorological conditions. Chemical transport models such as Community Multiscale Air Quality (CMAQ) and Weather Research and Forecasting Modeling with Chemistry (WRF-Chem) provide fundamental insights into the behavior of atmospheric pollutants but face limitations in spatial resolution and prediction accuracy. Machine learning combined with low-cost sensors can improve prediction accuracy compared to numerical models, but uneven sensor distribution may result in biased prediction results. To address these limitations, this study introduces an innovative framework that leverages a quadtree space division to refine the numerical simulations of air pollutants into higher spatial resolutions by dynamically adjusting grid sizes based on the distribution of increased sensor deployments. Using a variable-resolution grid, the study first aggregates variables into a grid and then calculates spatial dependence based on grid cells. The deep-learning-based Time Series Transformer model is then used to perform detailed temporal predictions of PM2.5 concentration across the entire grid. The spatial and temporal modeling framework is designed to use the past 48 h' aggregated data, including meteorological variables from WRF-Chem numerical simulation and PM2.5 concentrations from ground monitoring stations and sensors, to predict future 48-h’ PM2.5 concentrations and tested across entire grids within Los Angeles County from January to June 2020. The results show that the performance of variable resolution grids is better than that of fixed grids regarding four metrics: minimum resolution, overall spatiotemporal RMSE, average bias, and prediction coverage. Among them, the variable resolution grids with up to 4 sensors showed the best performance, with an overall spatiotemporal RMSE of 1.12 μg/m3, which was about 40% lower than 2-sensor and 8-sensor variable resolution grids, a significant reduction of 55% compared to a 3 km fixed grid configuration. Future work will involve additional relevant factors such as emission sources and sinks to improve prediction accuracy.

TransRUL: A Transformer-Based Multihead Attention Model for Enhanced Prediction of Battery Remaining Useful Life

The efficient operation of power-electronic-based systems heavily relies on the reliability and longevity of battery-powered systems. An accurate prediction of the remaining useful life (RUL) of batteries is essential for their effective maintenance, reliability, and safety. However, traditional RUL prediction methods and deep learning-based approaches face challenges in managing battery degradation processes, such as achieving robust prediction performance, to ensure scalability and computational efficiency. There is a need to develop adaptable models that can generalize across different battery types that operate in diverse operational environments. To solve these issues, this research work proposes a TransRUL model to enhance battery RUL prediction. The proposed model incorporates advanced approaches of a time series transformer using a dual encoder with integration positional encoding and multi-head attention. This research utilized data collected by the Centre for Advanced Life Cycle Engineering (CALCE) on CS_2-type lithium-ion batteries that spanned four groups that used a sliding window technique to generate features and labels. The experimental results demonstrate that TransRUL obtained superior performance as compared with other methods in terms of the following evaluation metrics: mean absolute error (MAE), root-mean-squared error (RMSE), and R2 values. The efficient computational power of the TransRUL model will facilitate the real-time prediction of the RUL, which is vital for power-electronic-based appliances. This research highlights the potential of the TransRUL model, which significantly enhances the accuracy of battery RUL prediction and additionally improves the management and control of battery-based systems.

Leak detection in water distribution networks based on deep learning and kriging interpolation method

ABSTRACT The burgeoning growth of urban areas has escalated the necessity for efficient and precise leak detection in water distribution networks. Automatic detection methods based on deep learning are a state-of-the-art research topic. In this paper, a methodology that combines deep learning and data imaging is proposed. The framework employs pressure monitoring data and is anchored on the following three pillars: (1) the generation of a comprehensive dataset, encompassing one year of leak-free demand data derived from Fourier Series analysis and monitoring pressure under normal and leak conditions, (2) the transformation of pressure time series into images using kriging interpolation, (3) establishing convolution neural network (CNN) and evaluating its performance of abnormal identification. The effectiveness of the proposed methodology is assessed in different image sets under various leak conditions. The findings reveal that this method meets dependable and effective outputs for leak detection, with the deep learning model achieving a high true positive rate (TPR) of 98% and an area under the curve (AUC) of 94%. This study provides invaluable information for strategic action planning and the enhancement of water loss management protocols, especially in situations where water utilities and regulatory authorities grappling with limited budgets and diminishing revenues.

A transformer-based forecasting model for [formula omitted] index and its application study on the Chinese Langfang dataset

The 10.7 cm solar radio flux (F10.7) is a key indicator of solar activity. Accurately forecasting of F10.7 is crucial for reducing the impact of solar activity on fields such as radio communication, navigation, and satellite communication. In this work, we present a novel channel-independent patch time series Transformer (PatchTST) for F10.7 forecasting. This is the first time that the PatchTST model is applied to F10.7 forecasting. We construct the F10.7 dataset, which is measured by the Dominion Radio Astrophysical Observatory (DRAO) in Canada. We compare the performance of PatchTST, N-Beats, BiGRU, and CNN-BiGRU on DRAO data. The root mean squared error (RMSE), mean absolute percentage error (MAPE), and correlation coefficient (R) of our PatchTST model are 4.731, 2.351%, and 0.986, respectively, which outperforms those of the other models when the prediction length is 1 day. Especially in mid-term forecasting, the PatchTST model performs much better than those of the other models. We make uncertainty analyses on these models, and the PatchTST model exhibits superior adaptability to model uncertainty compared to the N-Beats, BiGRU, and CNN-BiGRU. The PatchTST model shows a 62.9% improvement in mean error (ME) and a 40.5% improvement in standard mean error (STDE) compared to the benchmark data provided by Space Environment Technologies (SET). This work also shows that our PatchTST model generalizes well by applying it to other F10.7 observational data originating from Long and Short-band Solar Precision Flux Radiotelescope (L&S) in China.

Quantitative Changes in the Publishing of Lithuanian Scientific Journals in 2015–2022: Public and Private Sectors’ Adaptation to the Changing Conditions

The article analyses the quantitative changes in the Lithuanian scientific journal publishing in 2015-2022. The study reveals the problems of public bibliographic accounting and terminology in order to comprehensively analyse the public and private sector published print and electronic scientific journals. The gaps and uncertainties identified allow for the improvement of official publication records in the future, avoiding ambiguities in the assessment of the publication volumes of scientific journals. Science policy makers and implementers should take the initiative to standardise the definition of scientific journals, as this is the only way to achieve a clearer approach to this form of science communication. Although the annual volume of scientific journal publications has shrunk by 18.1% (from 265 (2015) to 217 (2022) publications), the decline itself has not been sharp. The decline has been stabilised by business involvement, changes in the evaluation of colleges, and the transformation of professional journals, book series and conference proceedings into scientific journals. Flexibility has been provided by electronic publishing instruments, which have reduced the dependence on external financial sources. The public publishing sector has tended to reduce the number of scientific journals in all scientific fields. This was mainly due to the reorganisation of the higher education network, the qualitative requirements for scientific articles and the reorganisation of scientific collectives. The publishing business has seen the greatest potential in the fields of technology and the medical and health sciences, and as a result, the number of journals published in these fields has not declined and continues to grow. Despite the private sector's growth trends, 50% of journals have 30 or more years of publication. This forms the core of Lithuanian scientific journals, which public sector publishers are striving to preserve and integrate into the international scientific environment.

An Improved Control Method of DC Voltage for Series Hybrid Active Power Filter

DC voltage is one of the important parameters of active power filters. Since the series hybrid active power filter does not withstand the fundamental voltage, it cannot absorb energy from the power grid, making it too difficult to control its DC voltage. In order to solve the DC voltage control problem, an improved control method for the series hybrid active power filter with magnetic flux compensation is proposed in this paper. In this improved method, the fundamental magnetic flux compensation coefficient is equal to 1, meeting the condition of fundamental magnetic flux compensation. This improved method does not rely on the precise detection and phase-locking of the fundamental voltage at the port of the series transformer. A phase-fixed active current component is generated directly by the inverter, enabling the active power filter (APF) to absorb active power from the power grid. Consequently, the active power absorbed by the APF from the power grid has a linear relationship with the active current component. Both simulation and experimentation verified the correctness and effectiveness of this proposed method.

Advancing predictive maintenance: a deep learning approach to sensor and event-log data fusion

PurposeThis study aims to introduce an innovative approach to predictive maintenance by integrating time-series sensor data with event logs, leveraging the synergistic potential of deep learning models. The primary goal is to enhance the accuracy of equipment failure predictions, thereby minimizing operational downtime.Design/methodology/approachThe methodology uses a dual-model architecture, combining the patch time series transformer (PatchTST) model for analyzing time-series sensor data and bidirectional encoder representations from transformers for processing textual event log data. Two distinct fusion strategies, namely, early and late fusion, are explored to integrate these data sources effectively. The early fusion approach merges data at the initial stages of processing, while late fusion combines model outputs toward the end. This research conducts thorough experiments using real-world data from wind turbines to validate the approach.FindingsThe results demonstrate a significant improvement in fault prediction accuracy, with early fusion strategies outperforming traditional methods by 2.6% to 16.9%. Late fusion strategies, while more stable, underscore the benefit of integrating diverse data types for predictive maintenance. The study provides empirical evidence of the superiority of the fusion-based methodology over singular data source approaches.Originality/valueThis research is distinguished by its novel fusion-based approach to predictive maintenance, marking a departure from conventional single-source data analysis methods. By incorporating both time-series sensor data and textual event logs, the study unveils a comprehensive and effective strategy for fault prediction, paving the way for future advancements in the field.

Inventory Prediction Using a Modified Multi-Dimensional Collaborative Wrapped Bi-Directional Long Short-Term Memory Model

Inventory prediction is concerned with the forecasting of future demand for products in order to optimize inventory levels and supply chain management. The challenges include demand volatility, data quality, multi-dimensional interactions, lead time variability, seasonal trends, and dynamic pricing. Nevertheless, these models suffer from numerous shortcomings, and in this research, we propose a new model, MMCW-BiLSTM (modified multi-dimensional collaboratively wrapped BiLSTM), for inventory prediction. The MMCW-BiLSTM model reflects a considerable leap in inventory forecasting by combining a number of components in order to consider intricate temporal dependencies and incorporate feature interactions. The MMCW-BiLSTM makes use of BiLSTM layers, collaborative attention mechanisms, and a multi-dimensional attention approach to learn from augmented datasets consisting of the original features and the extracted time series data. Moreover, adding a Taylor series transformation allows for a more precise description of the features in the model, thus improving the prediction precision. The results show that the models make the least mistakes when they use the AV demand forecasting dataset, with MAE values of 1.75, MAPE values of 2.89, MSE values of 6.76, and RMSE values of 2.6. Similarly, when utilizing the product demand dataset, the model also achieves the lowest error values for these metrics at 1.97, 3.91, 8.76, and 2.96. Likewise, when utilizing the dairy goods sales dataset, the model also achieves the lowest error values for these metrics at 2.54, 3.69, 10.39, and 3.22.

Insulation Design of High Voltage and Large Capacity Power Supply Transformer

Abstract At present, the highest voltage level of DC circuit breakers in the world is 535kV. With the construction of 800kV multi-terminal flexible DC project, the demand for DC circuit breaker with higher voltage level is put forward. The voltage level of valve base power supply transformer as the power supply component of the DC circuit breaker is increased to 800kV. Single-section pendulum structure, single-section 100kV and 50kV schemes used at 535kV cannot meet the demand. An 800kV valve base power supply transformer scheme of four 200kV gas-insulated transformers in series is proposed. Its insulation design is also carried out. Firstly, the topology of hybrid DC circuit breaker power supply system and the structure of power supply transformer were described. Secondly, the design of power supply transformer insulation was carried out. Finally, the power supply transformer insulation was tested and verified. The results show that: SF6 power supply transformer internal insulation withstand voltage needs to be multiplied by the multiplication factor. The electric field strength limit of SF6 obtained by experiments is more than 1.5 times of the empirical design value. The internal field strength design of the 200kV power supply transformer takes into account the operating and seismic conditions, and the maximum electric field strength is 15kV/mm and 16.2kV/mm respectively. The curvature radius of the top shielding ring needs to be larger than that of other layers, after optimization, the maximum field strength is 2.68kV /mm when the curvature radius is 300mm. The internal insulation design of 200kV power supply transformer according to lightning impulse. The simulated and measured values of non-uniform voltage coefficient of 800kV power supply transformer under lightning impulse are 1.048 and 1.0488 respectively, in which the voltage equalizing capacitance is 3000pF. The insulation tests of 200kV and 800kV power supply transformer were all passed, and it can be operated safely. Effectively support the development of UHV DC circuit breaker.