Proposed Long Short-term Memory Model Research Articles

State of health estimation embedded with hardware accelerator based on long short-term memory combined with Bayesian optimization considering extracted health indicator in charging conditions

Recently, the importance of renewable energy has increased as fossil fuel use is regulated worldwide to protect the environment. Accordingly, the use of electric vehicles (EVs) that use a battery as a power source is increasing and estimating the battery state for an efficient and safe operation of EV is a very significant task. This study proposes a method to estimate the state-of-health (SOH) in which batteries are a major state indicator. A health indicator (HI) capable of determining the state of the battery is derived from the charging part by considering the stopping situation in which there is a slight variation in the battery characteristics of the EV. As an integrity indicator, the partial capacity according to the voltage range set by the user can be considered, and the correlation between the SOH and HI is analyzed for validity verification. To learn the derived HI, a long short-term memory (LSTM) model, a type of deep-neural network, is designed, and model tuning was performed through Bayesian optimization. The absolute error of SOH estimation for the proposed LSTM model is 1.1 %, compared and validated against recurrent neural network (RNN) (1.7 %) and gated recurrent unit (GRU) (1.85 %). After validating the performance of the LSTM algorithm, an analysis of SOH estimation error (1 %) based on LSTM was conducted following pack-level experiments to assess scalability. Finally, for evaluation of operation within a hardware accelerator (HA) for integration into EVs, NVIDIA Jetson Nano was utilized. Upon mounting the algorithm on Jetson Nano, the mean absolute error (MAE) was 0.6581, mean absolute percentage error (MAPE) was 0.7209, and root mean square error (RMSE) was 0.7133, demonstrating excellent estimation performance.

Optimising air quality prediction in smart cities with hybrid particle swarm optimization‐long‐short term memory‐recurrent neural network model

AbstractIn smart cities, air pollution is a critical issue that affects individual health and harms the environment. The air pollution prediction can supply important information to all relevant parties to take appropriate initiatives. Air quality prediction is a hot area of research. The existing research encounters several challenges that is, poor accuracy and incorrect real‐time updates. This research presents a hybrid model based on long‐short term memory (LSTM), recurrent neural network (RNN), and Curiosity‐based Motivation method. The proposed model extracts a feature set from the training dataset using an RNN layer and achieves sequencing learning by applying an LSTM layer. Also, to deal with the overfitting issues in LSTM, the proposed model utilises a dropout strategy. In the proposed model, input and recurrent connections can be dropped from activation and weight updates using the dropout regularisation approach, and it utilises a Curiosity‐based Motivation model to construct a novel motivational model, which helps in the reconstruction of long short‐term memory recurrent neural network. To minimise the prediction error, particle swarm optimisation is implemented to optimise the LSTM neural network's weights. The authors utilise an online Air Pollution Monitoring dataset from Salt Lake City, USA with five air quality indicators for comparison, that is, SO2, CO, O3, and NO2, to predict air quality. The proposed model is compared with existing Gradient Boosted Tree Regression, Existing LSTM, and Support Vector Machine based Regression Model. Experimental analysis shows that the proposed method has 0.0184 (Root Mean Square Error (RMSE)), 0.0082 (Mean Absolute Error), 2002*109 (Mean Absolute Percentage Error), and 0.122 (R2‐Score). The experimental findings demonstrate that the proposed LSTM model had RMSE performance in the prescribed dataset and statistically significant superior outcomes compared to existing methods.

Construction of a recurrent neural network-based electrical load forecasting model for a 110 kV substation: a case study in the Western Region of The Republic of Kazakhstan

Construction of a recurrent neural network-based electrical load forecasting model for a 110 kV substation: a case study in the Western Region of The Republic of Kazakhstan

LSTM-Powered COVID-19 prediction in central Thailand incorporating meteorological and particulate matter data with a multi-feature selection approach

The COVID-19 pandemic has significantly impacted public health and necessitated urgent actions to mitigate its spread. Monitoring and predicting the outbreak's progression have become vital to devise effective strategies and allocate resources efficiently. This study presents a novel approach utilizing Multivariate Long Short-Term Memory (LSTM) to analyze and predict COVID-19 trends in Central Thailand, particularly emphasizing the multi-feature selection process. To consider a comprehensive view of the pandemic's dynamics, our research dataset encompasses epidemiological, meteorological, and particulate matter features, which were gathered from reliable sources. We propose a multi-feature selection technique to identify the most relevant and influential features that significantly impact the spread of COVID-19 in the region to enhance the model's performance. Our results highlight that relative humidity is the key factor driving COVID-19 transmission in Central Thailand. The proposed multi-feature selection technique significantly improves the model's accuracy, ensuring that only the most informative variables contribute to the predictions, avoiding the potential noise or redundancy from less relevant features. The proposed LSTM model demonstrates its capability to forecast COVID-19 cases, facilitating informed decision-making for public health authorities and policymakers.

A coupled data-physics computational framework for temperature, residual stress, and distortion modeling in autoclave process of composite materials

It is challenging to obtain the full-field temperature profile during autoclave processes to control the temperature uniformity and minimize the residual stress and distortion of cured composite structures. This paper proposes a coupled data-physics computational framework for full-field temperature reconstruction and the subsequent residual stress/distortion modeling by using limited monitoring temperature data. Firstly, a Long Short-Term Memory (LSTM) model is developed for full field temperature reconstruction. In this LSTM model, a cross-recombination method is proposed to maximize the value of monitored temperature data. The method effectively cuts through the bottleneck of neural network training with limited labeled data. The LSTM model’s prediction stability is enhanced based on the mean-teacher and ensemble learning strategy. To train and validate the proposed method, we perform experiments using an autoclave at the National Institute for Aviation Research (NIAR). The LSTM model’s accuracy is assessed by comparing its predicted results with the thermocouple (TC) data from measurements and high-fidelity simulation data from computational fluid dynamics (CFD). The study shows that the proposed LSTM model can effectively reconstruct the full-field temperature using limited monitoring data and significantly improve accuracy and efficiency compared with the CFD-based counterpart. Then, we create a coupled data-physics computational framework by embedding the data-driven LSTM model into a physics-based thermo-mechanical finite element model to predict residual stress and distortion. The simulation results show that the coupled data-physics framework provides an effective way for process-to-performance modeling and simulation of autoclave curing processes.

Diabetic Drug ontology mapping for individual Diabetic Person and predict Insulin Dosage on Daily Basis

Managing diabetes requires a personalized treatment plan based on the patient's health records and drug usage effectively. In this paper, we have proposed accurate prediction of insulin dosage for diabetic patients daily by using ontology mapping with deep learning recurrent neural network models such as long short-term memory (LSTM), Gated recurrent unit( GRU), and Bidirectional long short-term memory(BiLSTM). The patient's information is arranged in a structured format. The structured ontology data provides a meaningful connection between patients' personal information, physical activity, insulin intake, carbohydrate intake and glucose level. The structured data is beneficial to take decisions effectively to recommend the insulin dosage level daily to avoid a sudden decrease in sugar level. The proposed LSTM model outperforms well on insulin prediction compared to GRU and BiLSTM models. The proposed model is very effective for diabetic personal care and improves patient outcomes and quality of life.

Approach and Landing Energy Prediction Based on a Long Short-Term Memory Model

The statistical analysis of civil aircraft accidents reveals that the highest incidence of mishaps occurs during the approach and landing stages. Predominantly, these accidents are marked by abnormal energy states, leading to critical situations like stalling and heavy landings. Therefore, it is of great significance to accurately predict the aircraft energy state in the approach and landing stages to ensure a safe landing. In this study, a deep learning method based on time sequence data for the prediction of the aircraft approach and landing energy states is proposed. Firstly, by conducting an extensive overview of the existing literature, three characteristic parameters of altitude, velocity, and glide angle were selected as the indicators to characterize the energy state. Following this, a semi-physical simulation platform for a certain type of aircraft was developed. The approach and landing experiments were carried out with different throttle sizes and flap deflection under different wind speeds and wind directions. Then, a deep learning prediction model based on Long Short-Term Memory (LSTM) was established based on the experimental data to predict the energy state indicators during the approach and landing phases. Finally, the established LSTM model underwent rigorous training and testing under different strategies, and a comparative analysis was carried out. The results demonstrated that the proposed LSTM model exhibited high accuracy and a strong generalization ability in predicting energy states during the approach and landing phases. These results offer a theoretical basis for designing energy early warning systems and formulating the relevant flight control laws in the approach and landing stages.

Daily average load demand forecasting using LSTM model based on historical load trends

AbstractLoad demand forecasting is very important for the management, designing and analysis of an electrical grid system. Load forecasting has progressively become a crucial component of the energy management system with the growth of the smart micro grid. This study presents a new framework to long term load forecasting in the world of electricity power with the help of historical load trends. The main objective of this research work is estimating monthly electricity demand of an Indian state Chhattisgarh, in terms of per day average load demand using a machine learning model—Long Short‐Term Memory (LSTM). This framework considers average of each day load demand for every month of years 2018–2022 and forecasted per day average load demand for each month of the year 2023. Furthermore, the predicting accuracy is evaluated for training and testing phase, in terms of error metrics like Mean Absolute Error (MAE) and Mean Absolute Percentage Error (MAPE). The MAPE values under the training and testing phase are in the range of 0.010%–0.652% and 0.378%–10.54%, respectively. A comparative study of LSTM model with Artificial Neural Network (ANN) model indicates the proposed LSTM model is more accurate and can be applied for real time load demand forecasting.

Wind power deviation charge reduction using long short term memory network

High penetration of variable generation like wind in modern power systems results in frequent load-generation imbalances. Additional spinning reserves and balancing services are deployed to reduce or avoid such imbalances. However, these additional balancing services result in increased total operation costs and it is generally shifted to Wind Power Generation Companies (WPGC) as Deviation Charges (DCs). The high DCs diminish the profitability of WPGCs. DCs are directly proportional to forecasting errors and can be reduced by enhancing forecasting accuracy. This paper proposes a Long Short-Term Memory (LSTM) network to reduce DCs. LSTM is a special kind of recurrent neural network and it is capable of learning long-term and short-term dependencies effectively. Therefore, the LSTM network can produce minimal error wind power forecasts and this helps to reduce DCs. The performance of the proposed LSTM network is realized using the real system data. Results obtained from the proposed LSTM model are compared with various forecasting models like ARIMA, ANN, and SVR. The comparison shows that the proposed LSTM model shows the least total DCs on most of the days followed by ANN, SVR, and ARIMA. The proposed LSTM, ANN, SVR, and ARIMA models show an average total DC of 6194.25$, 8459.32$, 15822.22$, and 38269.89$ respectively. Thus, the proposed LSTM model is capable of reducing 517.82%, 155.43%, and 36.56% DCs from ARIMA, SVR, and ANN models respectively, and helps the WPGCs to enhance their profitability.

Detection of Autism Spectrum Disorder (ASD) Symptoms using LSTM Model

Autistic children will often exhibit certain behaviors that are unique to them and that are not typical of neurotypical children. Parents will become familiar with these patterns over time and will be able to use this knowledge to answer questions about their child's behavior. Deep learning models are very useful to solve critical problems in the healthcare domain. Detection of ASD at the early age of a child is a challenging task. Recent research reveals that there is an increasing trend of ASD among children. Communication, eye contact, social behavior, and education are very poor for those who suffer from ASD. The proposed research work has been done to detect ASD symptoms in a child. Data has been collected from the various autism groups from social sites and organizations that are working on special children. A Deep learning model like the Long-Short Term Memory (LSTM) model has been used to detect the sentiment of parents’ dialog. LSTM is the most popular deep learning model that can able to solve complex natural language problems. The proposed LSTM model has been trained with prepared data and accuracy is 97% according to the prepared data.

DECODE: Data-driven energy consumption prediction leveraging historical data and environmental factors in buildings

Energy prediction in buildings plays a crucial role in effective energy management. Precise predictions are essential for achieving optimal energy consumption and distribution within the grid. This paper introduces a Long Short-Term Memory (LSTM) model designed to forecast building energy consumption using historical energy data, occupancy patterns, and weather conditions. The LSTM model provides accurate short, medium, and long-term energy predictions for residential and commercial buildings compared to existing prediction models. We compare our LSTM model with established prediction methods, including linear regression, decision trees, and random forest. Encouragingly, the proposed LSTM model emerges as the superior performer across all metrics. It demonstrates exceptional prediction accuracy, boasting the highest R2 score of 0.97 and the most favorable mean absolute error (MAE) of 0.007. An additional advantage of our developed model is its capacity to achieve efficient energy consumption forecasts even when trained on a limited dataset. We address concerns about overfitting (variance) and underfitting (bias) through rigorous training and evaluation on real-world data. In summary, our research contributes to energy prediction by offering a robust LSTM model that outperforms alternative methods and operates with remarkable efficiency, generalizability, and reliability.

Enhancing Solar Radiation Forecasting in Diverse Moroccan Climate Zones: A Comparative Study of Machine Learning Models with Sugeno Integral Aggregation

Hourly solar radiation (SR) forecasting is a vital stage in the efficient deployment of solar energy management systems. Single and hybrid machine learning (ML) models have been predominantly applied for precise hourly SR predictions based on the pattern recognition of historical heterogeneous weather data. However, the integration of ML models has not been fully investigated in terms of overcoming irregularities in weather data that may degrade the forecasting accuracy. This study investigated a strategy that highlights interactions that may exist between aggregated prediction values. In the first investigation stage, a comparative analysis was conducted utilizing three different ML models including support vector machine (SVM) regression, long short-term memory (LSTM), and multilayer artificial neural networks (MLANN) to provide insights into their relative strengths and weaknesses for SR forecasting. The comparison showed the proposed LSTM model had the greatest contribution to the overall prediction of six different SR profiles from numerous sites in Morocco. To validate the stability of the proposed LSTM, Taylor diagrams, violin plots, and Kruskal–Wallis (KW) tests were also utilized to determine the robustness of the model’s performance. Secondly, the analysis found coupling the models outputs with aggregation techniques can significantly improve the forecasting accuracy. Accordingly, a novel aggerated model that integrates the forecasting outputs of LSTM, SVM, MLANN with Sugeno λ-measure and Sugeno integral named (SLSM) was proposed. The proposed SLSM provides spatially and temporary interactions of information that are characterized by uncertainty, emphasizing the importance of the aggregation function in mitigating irregularities associated with SR data and achieving an hourly time scale forecasting accuracy with improvement of 11.7 W/m2.

Deep causal speech enhancement and recognition using efficient long-short term memory Recurrent Neural Network.

Long short-term memory (LSTM) has been effectively used to represent sequential data in recent years. However, LSTM still struggles with capturing the long-term temporal dependencies. In this paper, we propose an hourglass-shaped LSTM that is able to capture long-term temporal correlations by reducing the feature resolutions without data loss. We have used skip connections in non-adjacent layers to avoid gradient decay. In addition, an attention process is incorporated into skip connections to emphasize the essential spectral features and spectral regions. The proposed LSTM model is applied to speech enhancement and recognition applications. The proposed LSTM model uses no future information, resulting in a causal system suitable for real-time processing. The combined spectral feature sets are used to train the LSTM model for improved performance. Using the proposed model, the ideal ratio mask (IRM) is estimated as a training objective. The experimental evaluations using short-time objective intelligibility (STOI) and perceptual evaluation of speech quality (PESQ) have demonstrated that the proposed model with robust feature representation obtained higher speech intelligibility and perceptual quality. With the TIMIT, LibriSpeech, and VoiceBank datasets, the proposed model improved STOI by 16.21%, 16.41%, and 18.33% over noisy speech, whereas PESQ is improved by 31.1%, 32.9%, and 32%. In seen and unseen noisy situations, the proposed model outperformed existing deep neural networks (DNNs), including baseline LSTM, feedforward neural network (FDNN), convolutional neural network (CNN), and generative adversarial network (GAN). With the Kaldi toolkit for automated speech recognition (ASR), the proposed model significantly reduced the word error rates (WERs) and reached an average WER of 15.13% in noisy backgrounds.

Equatorial spread-F forecasting model with local factors using the long short-term memory network

The predictability of the nighttime equatorial spread-F (ESF) occurrences is essential to the ionospheric disturbance warning system. In this work, we propose ESF forecasting models using two deep learning techniques: artificial neural network (ANN) and long short-term memory (LSTM). The ANN and LSTM models are trained with the ionogram data from equinoctial months in 2008 to 2018 at Chumphon station (CPN), Thailand near the magnetic equator, where the ESF onset typically occurs, and they are tested with the ionogram data from 2019. These models are trained especially with new local input parameters such as vertical drift velocity of the F-layer height (Vd) and atmospheric gravity waves (AGW) collected at CPN station together with global parameters of solar and geomagnetic activity. We analyze the ESF forecasting models in terms of monthly probability, daily probability and occurrence, and diurnal predictions. The proposed LSTM model can achieve the 85.4% accuracy when the local parameters: Vd and AGW are utilized. The LSTM model outperforms the ANN, particularly in February, March, April, and October. The results show that the AGW parameter plays a significant role in improvements of the LSTM model during post-midnight. When compared to the IRI-2016 model, the proposed LSTM model can provide lower discrepancies from observational data.Graphical

Predicting Choices Driven by Emotional Stimuli Using EEG-Based Analysis and Deep Learning

Individual choices and preferences are important factors that impact decision making. Artificial intelligence can predict decisions by objectively detecting individual choices and preferences using natural language processing, computer vision, and machine learning. Brain–computer interfaces can measure emotional reactions and identify brain activity changes linked to positive or negative emotions, enabling more accurate prediction models. This research aims to build an individual choice prediction system using electroencephalography (EEG) signals from the Shanghai Jiao Tong University emotion and EEG dataset (SEED). Using EEG, we built different deep learning models, such as a convolutional neural network, long short-term memory (LSTM), and a hybrid model to predict choices driven by emotional stimuli. We also compared their performance with different classical classifiers, such as k-nearest neighbors, support vector machines, and logistic regression. We also utilized ensemble classifiers such as random forest, adaptive boosting, and extreme gradient boosting. We evaluated our proposed models and compared them with previous studies on SEED. Our proposed LSTM model achieved good results, with an accuracy of 96%.

An "All-Data-on-Hand" Deep Learning Model to Predict Hospitalization for Diabetic Ketoacidosis in Youth With Type 1 Diabetes: Development and Validation Study.

Although prior research has identified multiple risk factors for diabetic ketoacidosis (DKA), clinicians continue to lack clinic-ready models to predict dangerous and costly episodes of DKA. We asked whether we could apply deep learning, specifically the use of a long short-term memory (LSTM) model, to accurately predict the 180-day risk of DKA-related hospitalization for youth with type 1 diabetes (T1D). We aimed to describe the development of an LSTM model to predict the 180-day risk of DKA-related hospitalization for youth with T1D. We used 17 consecutive calendar quarters of clinical data (January 10, 2016, to March 18, 2020) for 1745 youths aged 8 to 18 years with T1D from a pediatric diabetes clinic network in the Midwestern United States. The input data included demographics, discrete clinical observations (laboratory results, vital signs, anthropometric measures, diagnosis, and procedure codes), medications, visit counts by type of encounter, number of historic DKA episodes, number of days since last DKA admission, patient-reported outcomes (answers to clinic intake questions), and data features derived from diabetes- and nondiabetes-related clinical notes via natural language processing. We trained the model using input data from quarters 1 to 7 (n=1377), validated it using input from quarters 3 to 9 in a partial out-of-sample (OOS-P; n=1505) cohort, and further validated it in a full out-of-sample (OOS-F; n=354) cohort with input from quarters 10 to 15. DKA admissions occurred at a rate of 5% per 180-days in both out-of-sample cohorts. In the OOS-P and OOS-F cohorts, the median age was 13.7 (IQR 11.3-15.8) years and 13.1 (IQR 10.7-15.5) years; median glycated hemoglobin levels at enrollment were 8.6% (IQR 7.6%-9.8%) and 8.1% (IQR 6.9%-9.5%); recall was 33% (26/80) and 50% (9/18) for the top-ranked 5% of youth with T1D; and 14.15% (213/1505) and 12.7% (45/354) had prior DKA admissions (after the T1D diagnosis), respectively. For lists rank ordered by the probability of hospitalization, precision increased from 33% to 56% to 100% for positions 1 to 80, 1 to 25, and 1 to 10 in the OOS-P cohort and from 50% to 60% to 80% for positions 1 to 18, 1 to 10, and 1 to 5 in the OOS-F cohort, respectively. The proposed LSTM model for predicting 180-day DKA-related hospitalization was valid in this sample. Future research should evaluate model validity in multiple populations and settings to account for health inequities that may be present in different segments of the population (eg, racially or socioeconomically diverse cohorts). Rank ordering youth by probability of DKA-related hospitalization will allow clinics to identify the most at-risk youth. The clinical implication of this is that clinics may then create and evaluate novel preventive interventions based on available resources.

Predicting Commercial Building Energy Consumption Using a Multivariate Multilayered Long-Short Term Memory Time-Series Model

The global demand for energy has been steadily increasing due to population growth, urbanization, and industrialization. Numerous researchers worldwide are striving to create precise forecasting models for predicting energy consumption to manage supply and demand effectively. In this research, a time-series forecasting model based on multivariate multilayered long short-term memory (LSTM) is proposed for forecasting energy consumption and tested using data obtained from commercial buildings in Melbourne, Australia: the Advanced Technologies Center, Advanced Manufacturing and Design Center, and Knox Innovation, Opportunity, and Sustainability Center buildings. This research specifically identifies the best forecasting method for subtropical conditions and evaluates its performance by comparing it with the most commonly used methods at present, including LSTM, bidirectional LSTM, and linear regression. The proposed multivariate, multilayered LSTM model was assessed by comparing mean average error (MAE), root-mean-square error (RMSE), and mean absolute percentage error (MAPE) values with and without labeled time. Results indicate that the proposed model exhibits optimal performance with improved precision and accuracy. Specifically, the proposed LSTM model achieved a decrease in MAE of 30%, RMSE of 25%, and MAPE of 20% compared with the LSTM method. Moreover, it outperformed the bidirectional LSTM method with a reduction in MAE of 10%, RMSE of 20%, and MAPE of 18%. Furthermore, the proposed model surpassed linear regression with a decrease in MAE by 2%, RMSE by 7%, and MAPE by 10%.These findings highlight the significant performance increase achieved by the proposed multivariate multilayered LSTM model in energy consumption forecasting.

LSTM Network for the Oxygen Concentration Modeling of a Wastewater Treatment Plant

The activated sludge process is a well-known method used to treat municipal and industrial wastewater. In this complex process, the oxygen concentration in the reactors plays a key role in the plant efficiency. This paper proposes the use of a Long Short-Term Memory (LSTM) network to identify an input–output model suitable for the design of an oxygen concentration controller. The model is identified from easily accessible measures collected from a real plant. This dataset covers almost a month of data collected from the plant. The performances achieved with the proposed LSTM model are compared with those obtained with a standard AutoRegressive model with eXogenous input (ARX). Both models capture the oscillation frequencies and the overall behavior (ARX Pearson correlation coefficient ρ = 0.833 , LSTM ρ = 0.921), but, while the ARX model fails to reach the correct amplitude (index of fitting FIT = 41.20%), the LSTM presents satisfactory performance (FIT = 60.56%).

Deep learning approach for one-hour ahead forecasting of weather data

ABSTRACT Weather is made up of multiple parameters, including solar radiation (SR), atmospheric pressure (AP), soil temperature (ST), atmospheric temperature (AT), wind speed (WS), relative humidity (RH), and sunshine duration (SD). These factors are also crucial for the renewable energy sector, solar simulation, agriculture, air pollution, water supply and distribution, avalanche warning, forestry, and town and regional planning. A deep learning method based on a neural network with Long Short-Term Memory (LSTM) was employed in this investigation for one-hour-ahead weather data forecasting. The ability of the LSTM model was compared with the Adaptive Neuro-Fuzzy Inference System (ANFIS) with that of the fuzzy c-means (FCM), Autoregressive Integrated Moving Average (ARIMA) model, and the Autoregressive Moving Average (ARMA) model. Mean absolute error (MAE), correlation coefficient (R), root means square error (RMSE), average bias, Nash – Sutcliffe efficiency coefficient (NSE), and mean absolute percentage error (MAPE) were selected as evaluation criteria. Results indicated that the proposed LSTM model presented good enough results compared to other used methods. 7 different types of meteorological data from a total of 4 years (35040 hours) were divided into 25% test data and 75% training data for the models. The best result was obtained for the hourly ST estimation of Adana province using the LSTM method, the MAE, RMSE, R, bias, NSE, and MAPE values were computed as 0.016°C, 0.078°C, 0.9999, −0.00018°C, 0.0805%, and 0.9999, respectively. On the other hand, the worst result was obtained for the hourly SD for Mardin province when ARIMA was used, and the statistical measures were derived as 0.128 hours for MAE, 0.215 hours for RMSE, 0.8851 for R, 0.00091 hours for bias, and 0.7657 for NSE. In this regard, it is demonstrated that the LSTM technique outperformed the other models in terms of all-weather data estimates and delivered highly sensitive outcomes.

Modeling Nonlinear Aeroelastic Forces for Bridge Decks with Various Leading Edges Using LSTM Networks

With the rapid increase in bridge spans, the mitigation of risk to flutter (aeroelastic instability) is of critical importance in the design of long-span bridges, especially considering the more frequent intense hurricanes under climate change. Although the strong nonlinearities of the aeroelastic (self-excited) forces in wind–bridge interactions can be well captured through either numerical simulations or experimental tests, both are expensive and time consuming. Hence, it is important to develop an efficient reduced-order model for the simulations of nonlinear aeroelastic forces on the bridge decks. This study proposes a reduced-order model based on the long short-term memory (LSTM) network to simulate the nonlinear aeroelastic forces on bridge decks with various leading edges, and thus rapidly predict the corresponding post-flutter behaviors of long-span bridges. To generate the training datasets, computational fluid dynamics (CFD) was employed to simulate the nonlinear aeroelasticities of bridge decks with a wide range of leading-edge configurations and wind speeds. Trained on the high-fidelity CFD datasets, the LSTM network takes the motion of a bridge deck, leading-edge angles and wind speeds as inputs and outputs the nonlinear aeroelastic forces on the bridge decks. A hybrid loss function utilizing the prediction errors of both aeroelastic forces simulated by the LSTM network and the bridge deck responses calculated by the Newmark-β algorithm was introduced into the training process to improve the network performance. The prediction results of the trained LSTM model were compared with the CFD simulations, which demonstrated that the nonlinear aeroelastic forces of the bridge deck with various leading edges can be accurately and efficiently acquired by the proposed LSTM model.