Recurrent Unit Neural Network Research Articles

Intra-hour photovoltaic power point-interval prediction using a dual-view deep neural network

Photovoltaic (PV) power is intermittent, fluctuating, and uncertain due to meteorological factors, which makes it challenging to integrate PV power into electrical grids. Overcoming this problem requires an accurate and reliable PV power prediction model; thus, a dual-view deep neural network (Dv-DNN) is proposed. The model fuses a gate recurrent unit (GRU) neural network and a temporal convolutional network (TCN) in parallel, enabling end-to-end dual-view deep prediction for PV power. Using data from solar systems in the Yulara region of Australia, three sets of experiments—single-step, multi-step, and interval prediction—are conducted to explore the performance of two recurrent and convolutional deep neural networks in PV power prediction. For point prediction, the experimental results show that the Dv-DNN has better forecasting performance than other models as the prediction horizon increases. Regarding interval prediction, at confidence levels of 97.5%, 95%, 90%, and 85%, the method based on kernel density estimation (KDE) is superior to the method based on maximum likelihood estimation (MLE) according to the coverage width-based criterion (CWC).

Metal Futures Prices Prediction Based on LSTM-GRU Hybrid Model

Accurately predicting the prices of metal futures such as aluminum, copper, lead, nickel, tin, and zinc is crucial for investors, manufacturers, and policymakers due to their significant impact on economic activities and industrial processes. This paper presents a comprehensive study on the prediction of futures prices using a hybrid model combining Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) neural networks. The research utilizes historical data from 2014 to 2024, including daily open, high, low, close prices, trading volumes, and return. Data preprocessing involved interpolation for missing values and normalization using MinMaxScaler. The models performance was evaluated with various accuracy metrics, such as Root Mean Squared Error (RMSE), Mean Squared Error (MSE), Mean Absolute Error (MAE) ect, showing high accuracy for lead, tin, and zinc futures in both train and test datasets. However, the model performed well only on the train data for aluminum, copper, and nickel, indicating the need for more iterations, additional factors, or the consideration of unexpected events. The results highlight the efficacy of the LSTM-GRU hybrid model in futures price prediction, with significant implications for informed investment decisions. Future research should consider cross-market analysis and additional factors to enhance model accuracy further.

Risk identification and prediction model for continuous‐lane‐change vehicles considering driving style

Lane change risk identification and prediction research is important for traffic safety, development of advanced driver assistance systems, traffic flow optimization. Among them, vehicle continuous‐lane‐change (CLC) risk identification and prediction is weak, but this element is important for vehicle decision-making and path planning in multi-lane, non-adjacent lane situations. To compensate for the shortcomings of risk identification based on both deterministic and probabilistic methods and address the issues of poor measurement accuracy and efficiency caused by manual tuning, this study established a risk identification and prediction model based on the Bayesian global optimization (BO) gated recurrent unit (GRU) neural network (BO-GRU) and instantiated the model using the trajectory data of 129 vehicles. Overall, the results showed that the accuracy of identifying CLC risks and driving-style was 99.33% and 97.44%, respectively, while the overall average accuracy of predicting potential CLC risks based on the identification of risks and predicted trajectories was 97.13%. The root-mean-square errors of CLC predicted trajectory for different driving-styles were 0.2503, 0.2331, 0.2093, and 0.2706, respectively, which were 84%−88% lower compared to 1.7361 when driving-styles were not considered. The performance of the BO-GRU model is comparable to that of parameter-optimized LSTM and its variant models, which highlights the significance of parameter optimization for neural network models. Overall, the use of a BO-GRU-based model for identifying and predicting continuous lane change risks in vehicles can aid in making driving decisions and planning routes, decrease the risk of collisions, enhance traffic flow efficiency, and encourage safe driving behavior.

State of charge estimation for the lithium-ion battery based on fractional-order multi-dimensional Taylor network

To accurately estimate the state of charge (SOC) of the lithium-ion battery (LiB), a fractional-order multi-dimensional Taylor network (FMTN) model was proposed in the study. With two open datasets of LiBs, the performance of the FMTN model on the SOC estimation is evaluated and compared with multi-dimensional Taylor network (MTN), back propagation neural network (BPNN), and dual-stage attention gate recurrent unit neural network (DA-GRUNN) models. With the datasets of B0005 and B0006 LiBs, the SOC estimation accuracy of the FMTN-6 model is 34 %, 32 %, and 37 %, 45 % higher than that of the MTN-6 and BPNN-3 models, respectively. The estimation time consumptions of the FMTN-6 model are increased by 2.8 % and 3.9 % and reduced by 49.5 % and 52.8 % compared with that of the MTN-6 and BPNN-3 models, respectively. With the open dataset of real driving conditions, the root-mean-square error of the SOC estimation of the FMTN model is reduced by 48 % compared with that of the MTN model. Besides, with the datasets at 25 °C and 45 °C, the SOC estimation accuracy of the FMTN model is improved by 17 % and 18 % compared with that of the DA-GRUNN model, respectively.

DeepTimeNet: A novel architecture for precise surface temperature estimation of lithium-ion batteries across diverse ambient conditions

With the growing demand for battery-powered devices and electric vehicles, the need for improved battery performance and safety is paramount. A key determinant of battery health is the accurate monitoring of surface temperature (ST). Conventional ST estimation often depends on direct sensor measurements, which may not be cost-effective and can impact system reliability. This paper presents DeepTimeNet, a novel approach leveraging deep learning (DL) architectures for sensorless ST prediction in lithium-ion batteries. DeepTimeNet combines Convolutional Neural Networks (CNN), ResNet blocks, Inception modules, Bidirectional LSTM, and GRU layers to precisely model the time-dependent behaviour of batteries. A comprehensive evaluation against traditional models, across temperatures ranging from -20 °C to 25 °C and under various driving profiles, including US06 and Urban Dynamometer Driving Schedule (UDDS), is conducted. DeepTimeNet’s performance is quantified by metrics such as mean absolute error (MAE), surpassing that of models like Gated Recurrent Unit-Recurrent Neural Network (GRU-RNN), Convolutional Neural Network-Long Short Term Memory Network (CNN-LSTM), and Long Short Term Memory Network (LSTM). The results demonstrate DeepTimeNet’s superior performance, with an RMSE of 0.0971, MSE of 0.0099, MAE of 0.0912, and MAXE of 0.3963, validating it as an advanced tool for enhancing the efficacy of battery management systems and underscoring its potential as a benchmark for future innovations.

An efficient neural network approach for laminated composite plates using refined zigzag theory

This paper presents an innovative methodology employing the One-dimensional Convolutional Gated Recurrent Unit neural network (1D-CGRU) algorithm for the analysis of laminated composites using the Refined Zigzag theory (RZT). The RZT methodology is utilized to assess laminated plate structures and generate essential data, forming the basis for training the 1D-CGRU model. The synergistic application of RZT and 1D-CGRU demonstrates exceptional global–local accuracy in predicting the mechanical behavior of laminated composite plates. For efficient data generation, RZT not only provides high precision, but also exhibits computational efficiency, making it suitable for finite element simulations with a C0-continuous kinematic approximation. Additionally, the 1D-CGRU model integrates the strengths of a One-dimensional Convolutional Neural Network (1D-CNN) for spatial feature extraction and dimensionality reduction, coupled with a Gated Recurrent Unit (GRU) network for discerning temporal relationships and mapping them to the target domain. Furthermore, quantitative accuracy measurements, including Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE), are used to validate the superior performance of the 1D-CGRU model compared to other surrogate models. For angle-ply and cross-ply composite laminates, the 1D-CGRU model achieves remarkable accuracy (98.15% and 99.45%, respectively) with low RMSE values. These results highlight the potential of the proposed framework to enhance predictive analysis for laminated composite structures, offering valuable insights for engineering applications and design optimizations.

A novel improved hybrid neural network for predicting heating load in airport building

In order to guarantee the efficient and energy-saving operation and control of airport heating systems, accurate heating load prediction is extremely critical. In view of the relatively little research on heating load prediction of airport buildings, besides, out of these, the fact that the actual characteristics of the airport itself and its production operations should be considered, resulting in inferior performance. Framed in this goal, Initially, the heating area is partitioned. Subsequently, a complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN) algorithm synthesizes it with fuzzy entropy (FE) to decompose the heating load, and the maximum information coefficient integrated with the distance correlation coefficient is utilized for the feature selection. After that, a multi-strategy improved artificial bee colony algorithm (MIABC) algorithm is used to optimize the radial basis function (RBF)neural network and the gated recurrent unit (GRU) neural network, respectively, and the two are combined to construct the hybrid prediction model of the heating load. Finally, actual heating load data is used to verify the prediction performance of the proposed model. Besides, the results show that the root mean square error (RMSE), mean absolute percentage error (MAPE), range of relative error (RRE), and mean bias error (MBE) of the proposed model are 0.465, 1.08 %, 11.42 % and 0.082, respectively, the coefficient of determination (R2) is 0.997, and the prediction accuracy reaches 98.92 %. Additionally, it has more robustness and generalization power, effectively overcoming the heating load's uncertainties and laying a foundation for controlling, monitoring, and managing the buildings' operation.

Performance prediction method of solid rocket motor based on the ground test

Abstract Solid rocket motor (SRM) is a core component of rockets. Considering the high cost of solid rocket motors and the high demand for transportation and storage, there is an urgent need to predict the performance of solid rocket motors before the ground ignition test to lower the cost of tests. Using data mining technology to predict various parameters of rocket motors with historical data has become a new development direction. In order to improve the data fitting effect and prediction accuracy, a novel prediction model was proposed, which tries to combine the Whale Optimization Algorithm (WOA) with the Error Correction Gated Recurrent Unit neural network (ECGRU) to correct the basic prediction performance of the Long Short-Term Memory neural network model (LSTM). Compared with traditional methods, experiments showed that the novel model was beneficial in improving the reliability of the prediction of the thrust and specific impulse.

A novel fatigue driving detection method based on whale optimization and Attention-enhanced GRU

Fatigue driving has emerged as the predominant causative factor for road traffic safety accidents. The fatigue driving detection method, derived from laboratory simulation data, faces challenges related to imbalanced data distribution and limited recognition accuracy in practical scenarios. In this study, we introduce a novel approach utilizing a gated recurrent neural network method, employing whale optimization algorithm for fatigue driving identification. Additionally, we incorporate an attention mechanism to enhance identification accuracy. Initially, this study focuses on the driver's operational behavior under authentic vehicular conditions. Subsequently, it employs wavelet energy entropy, scale entropy, and singular entropy analysis to extract the fatigue-related features from the driver's operational behavior. Subsequently, this study adopts the cross-validation recursive feature elimination method to derive the optimal fatigue feature index about operational behavior. To effectively capture long-range dependence relationships, this study employs the gated recurrent unit neural network method. Lastly, an attention mechanism is incorporated in this study to concentrate on pivotal features within the data sequence of driving behavior. It assigns greater weight to crucial information, mitigating information loss caused by the extended temporal sequence. Experimental results obtained from real vehicle data demonstrate that the proposed method achieves an accuracy of 89.84% in third-level fatigue driving detection, with an omission rate of 10.99%. These findings affirm the feasibility of the approach presented in this study.

Water-to-air unmanned aerial vehicle (UAV) based rolling shutter optical camera communication (OCC) system with gated recurrent unit neural network (GRU-NN)

Water-to-air unmanned aerial vehicle (UAV) based rolling shutter optical camera communication (OCC) system with gated recurrent unit neural network (GRU-NN)

Predicting Structural Temperature of Ancient City Walls: A Case Study Using Ambient Temperature and Deep Learning

ABSTRACT Accurate prediction of the structural temperature field is required for the performance assessment of engineering structures and for maintaining structural safety. However, there are complex time-lag effects between ambient temperature and structural temperature. It is difficult to accurately describe and predict the structural temperature using traditional linear and nonlinear fitting methods. Hence, this study uses the advantages of gated recurrent unit (GRU) neural networks in dealing with complex nonlinear mapping issues, further considering solar radiation and developing a structural temperature prediction framework using ambient temperature and the deep learning model. The framework is applied in structural temperature prediction for an ancient city wall in China. The solar altitude angle at the location of the city wall is calculated, and the input and output structures of the deep learning neural network model are optimized. In the long-term prediction of the structural temperature of the city wall, the root mean square error (RMSE) of the GRU model is less than 0.6°C, which solves the problem of low accuracy caused by time-lag effects.

Identification of switched gated recurrent unit neural networks with a generalized Gaussian distribution

Identification of switched gated recurrent unit neural networks with a generalized Gaussian distribution

A hybrid information fusion method for SINS/GNSS integrated navigation system utilizing GRU-aided AKF during GNSS outages

Abstract To enhance the performance of integrated inertial navigation system (INS) and global navigation satellite system (GNSS) during GNSS outages, this paper proposed a fusion positioning method based on predictive observation information and adaptive filter parameter. Combined with an adaptive Kalman filter (AKF) and a Gated Recurrent Unit neural network (NN) that directly relates the inertial measurement unit (IMU) output sequence to the error estimation, the hybrid information fusion system can provide effective corrections to compensate for horizontal position errors under the constraints of complex and dynamic vehicle movement data during GNSS outages. Meanwhile, the designed adaptive parameter of the integrated navigation filter can adjust the credibility of the state prediction section when the GNSS is reconnected, ensuring the system can switch rapidly between the INS/GNSS and INS/NN integrated modes. The performance of the proposed information fusion method has been experimentally validated using IMU and GNSS data collected in a vehicle navigation test conducted on a stretch of expressway. The comparison results indicate that the proposed algorithm has error suppression capabilities under various experimental constraints and demonstrates a degree of extendibility and reusability.

A Variational Mode Decomposition–Grey Wolf Optimizer–Gated Recurrent Unit Model for Forecasting Water Quality Parameters

Water is a critical resource globally, covering approximately 71% of the Earth’s surface. Employing analytical models to forecast water quality parameters based on historical data is a key strategy in the field of water quality monitoring and treatment. By using a forecasting model, potential changes in water quality can be understood over time. In this study, the gated recurrent unit (GRU) neural network was utilized to forecast dissolved oxygen levels following variational mode decomposition (VMD). The GRU neural network’s parameters were optimized using the grey wolf optimizer (GWO), leading to the development of a VMD–GWO–GRU model for forecasting water quality parameters. The results indicate that this model outperforms both the standalone GRU model and the GWO–GRU model in capturing key information related to water quality parameters. Additionally, it shows improved accuracy in forecasting medium to long-term water quality changes, resulting in reduced root mean square error (RMSE) and mean absolute percentage error (MAPE). The model demonstrates a significant improvement in the lag of forecasting water quality parameters, ultimately boosting forecasting accuracy. This approach can be applied effectively in both monitoring and forecasting water quality parameters, serving as a solid foundation for future water quality treatment strategies.

Enhanced SOC estimation in Lithium-ion Batteries using GRU Neural Networks with GSA

State of charge (SOC) is one of the most important metrics to evaluate the performance of the battery management system (BMS) in electric vehicles (EVs). Lithium-ion batteries have been utilized often for SOC estimates in EV applications because of its attractive characteristics such as extended lifespans, high voltages, energy, and capacities. However, dynamic charging and discharging profiles, material degradation, battery aging, chemical reaction, and temperature fluctuations would diverge the accuracy of SOC estimation. Deep learning has proven to become an efficient method for SOC estimation in EVs due to its strong computational capabilities, enhanced generalization performance, and excellent accuracy under dynamic loading profiles. Therefore, this study proposes the use of the gated recurrent unit neural network (GRUNN) for estimating SOC in lithium-ion batteries. The best hyperparameters of GRUNN that enable improving SOC accuracy are found using gravitational search optimization (GSA). To evaluate the functional sustainability of the suggested method, a full battery test bench model is created and accordingly, robustness of the proposed approach is verified under diverse EV drive cycles and aging impacts. The effectiveness of the proposed approach is evaluated by comparing with several existing approaches. The results demonstrate that the suggested method archives SOC error and RMSE in EV driving cycles and ageing effects below 5% and 2%, respectively. The excellent outcomes obtained by the proposed method would certainly enhance the battery charging and discharging profile and EV performance. GUB JOURNAL OF SCIENCE AND ENGINEERING, Vol 9(1), 2022 P 1-12

Analyzing sleep thermal comfort with an attention-based gated recurrent unit neural network

To provide a comfortable sleep environment, the air conditioning controller needs a feedback mechanism associated with sleep thermal comfort. A thermal comfort prediction system is therefore needed. However, there is a dearth of research in this area. Recent advancements in wearable sensor technology and deep learning provide innovative solutions for monitoring sleep comfort. In this study, we propose a novel attention-based gated recurrent unit neural network (AGRU) designed for the prediction of sleep thermal comfort. This model leverages data from air environment monitors and wearable sensors capable of detecting environmental variables and physiological signals. We conducted an eighty-day sleep experiment involving twenty subjects exposed to varying environmental conditions. The monitors and sensors recorded seven features, including air temperature, relative humidity, skin temperatures at four points, and pulse rate. The results obtained underline the practicality and efficacy of the deep learning model that draws on environmental and physiological signals, with an average accuracy, macro-precision, macro-recall and macro-F1-score of 86.99 %, 87.10 %, 86.98 % and 0.87, respectively. This research provides substantial support for the continued advancement of smart home technology and wearable technology in the field of sleep thermal comfort.

AIR POLLUTION PREDICTION USING MODIFIED GRU DEEP LEARNING MODEL

<p>In recent days air quality has been very poor in urban society due to a lot of industries, vehicles and so on. To overcome poor quality, an air quality prediction is much needed to manage an urban area and control pollution. In this article, deep learning (DL) namely a Modified Gated Recurrent Unit (GRU) neural network is used to forecast air pollution. Initially, a new activation function named Dual-Slope Leaky ReLU (DSL-ReLU) is introduced. The DSL-ReLU activation function includes two hyperparameters which control the slopes for positive and negative input values to achieve fine-tuned responses for varying data inputs. Then, the model parameters like learning rate, batch size, and dropout rates are tuned using the Spider Wasp Optimization (SWO) algorithm. The SWO algorithm is based on the female spider wasp’s hunting and nesting behaviour that is applied to solve the optimization issues. This SWO model is used to GRU model parameters and provides robustness in the predictive modelling of air quality monitoring. For validation, the Kaggle data set is used whereas the proposed optimized GRU model considerably improves the accuracy of air pollution forecasts</p>

Offshore wind power prediction based on two-stage hybrid modeling**

Accurate prediction of offshore wind power generation is essential for efficient power scheduling and grid integration. This study introduces an innovative hybrid forecasting approach to address variability in power output due to different climatic conditions and challenges in measuring offshore wind speeds. The approach consists of two primary phases. First, a gated recurrent unit neural network with an attentional mechanism and a quantile regression model forecast wind speeds between cut-in and rated wind speeds, capturing trends under distinct climatic conditions. Second, wind speeds at nine quantile points are used as inputs to a relevance vector machine model, optimized via a cuckoo search algorithm, to establish the relationship between wind speed and power output. An empirical evaluation on a European offshore wind power dataset validates the approach's effectiveness across various climatic conditions. The two-stage model demonstrates enhanced adaptability, offering more reliable power predictions than conventional methods. Results indicate this hybrid forecasting method is more accurate than traditional techniques, significantly improving offshore wind power prediction performance.

ResGRU: A deep learning approach for settlement prediction in CFRD based on the spatiotemporal feature fusion method

Settlement is one of the most critical safety evaluation indicators for concrete face rock-fill dams. Deformation measurement data exhibit spatiotemporal correlation due to similar environmental factors and material properties. However, the current research has focused mainly on single-point prediction models, which ignore the relationship between settlement values and various parts of dams. To address these issues, the spatiotemporal correlation of settlements across multiple measurement points is considered and a deep learning approach based on spatiotemporal feature fusion is proposed. First, a density peak clustering algorithm categorizes monitoring points characterized by analogous deformation patterns. Then, a deep learning model combining residual neural networks (ResNet) and gated recurrent unit neural networks (GRU), i.e. ResGRU, is established. In ResGRU, ResNet is used to map the input data’s features to more complex high-dimensional patterns through convolution and obtain deep nonlinear spatial features. GRU is used to capture the temporal dependencies of data to obtain temporal features. Finally, the spatiotemporal features are fused through fully connected layers to predict settlements. The model is validated and compared with deep learning models using monitoring data collected from an actual concrete face rock-fill dam. The proposed approach demonstrates superior performance with an average RMSE and MAE of 0.352 and 0.255, respectively, and an average R2 of 0.988. The results show that the ResGRU model generally outperforms the other models. This approach provides an efficient and convenient approach for the settlement analysis of concrete face rock-fill dams.

Dyna algorithm-based reinforcement learning energy management for fuel cell hybrid engineering vehicles

Fuel cell hybrid engineering vehicles utilize green energy with high energy conversion efficiency. Optimizing the utilization of hybrid energy storage systems is crucial for improving energy efficiency and promoting energy conservation. This study introduces an innovative energy management framework that employs reinforcement learning to regulate a fuel cell/battery hybrid engineering vehicle. This novel framework facilitates the system achieve optimal energy management, even when confronted with demanding and ever-changing operational scenarios. The proposed strategy integrates the concepts of the Dyna algorithm and the deep Q-network (DQN) algorithm. A working conditions predictive model is developed using the gated recurrent unit (GRU) neural network. The aim of this predictive model is to generate a virtual depiction of the environment and accelerate the process of model updating. The results from simulations and hardware-in-the-loop test illustrate that the predictive model exhibits effective prediction capabilities, enabling the development of an accurate environmental model even when limited sample data is available. The proposed strategy exhibits superior performance compared to the DQN and Q-Learning algorithms regarding adaptability, real-time optimization, training efficiency, convergence speed, and energy conservation. When interacting with the environment, our proposed learning algorithm exhibits a faster convergence rate. Moreover, this strategy achieves a fuel economy of 95 %, as determined by the dynamic programming-based method. The model-based deep reinforcement learning approach is practical and does not require extensive training. An added benefit of this energy management architecture is its use of a more accurate world model.