Dynamic Artificial Neural Network Model Research Articles

Wastewater-based epidemiology for COVID-19 using dynamic artificial neural networks

Global efforts in vaccination have led to a decrease in COVID-19 mortality but a high circulation of SARS-CoV-2 is still observed in several countries, resulting in some cases of severe lockdowns. In this sense, wastewater-based epidemiology remains a powerful tool for supporting regional health administrations in assessing risk levels and acting accordingly. In this work, a dynamic artificial neural network (DANN) has been developed for predicting the number of COVID-19 hospitalized patients in hospitals in Valladolid (Spain). This model takes as inputs a wastewater epidemiology indicator for COVID-19 (concentration of RNA from SARS-CoV-2 N1 gene reported from Valladolid Wastewater Treatment Plant), vaccination coverage, and past data of hospitalizations. The model considered both the instantaneous values of these variables and their historical evolution. Two study periods were selected (from May 2021 until September 2022 and from September 2022 to July 2023). During the first period, accurate predictions of hospitalizations (with an overall range between 6 and 171) were favored by the correlation of this indicator with N1 concentrations in wastewater (r = 0.43, p < 0.05), showing accurate forecasting for 1 day ahead and 5 days ahead. The second period's retraining strategy maintained the overall accuracy of the model despite lower hospitalizations. Furthermore, risk levels were assigned to each 1 day ahead prediction during the first and second periods, showing agreement with the level measured and reported by regional health authorities in 95 % and 93 % of cases, respectively. These results evidenced the potential of this novel DANN model for predicting COVID-19 hospitalizations based on SARS-CoV-2 wastewater concentrations at a regional scale. The model architecture herein developed can support regional health authorities in COVID-19 risk management based on wastewater-based epidemiology.

Event-Triggered $H_{\infty}$ Tracking Control for Dynamic Artificial Neural Network Models With Time-Varying Delays

Event-Triggered $H_{\infty}$ Tracking Control for Dynamic Artificial Neural Network Models With Time-Varying Delays

Online interpretable dynamic prediction models for postoperative delirium after cardiac surgery under cardiopulmonary bypass developed based on machine learning algorithms: A retrospective cohort study

ObjectivePostoperative delirium (POD) is strongly associated with poor early and long-term prognosis in cardiac surgery patients with cardiopulmonary bypass (CPB). This study aimed to develop dynamic prediction models for POD after cardiac surgery under CPB using machine learning (ML) algorithms. MethodsFrom July 2021 to June 2022, clinical data were collected from patients undergoing cardiac surgery under CPB at Nanjing First Hospital. A dataset from the same center (October 2022 to November 2022) was also used for temporal external validation. We used ML and deep learning to build models in the training set, optimized parameters in the test set, and finally validated the best model in the validation set. The SHapley Additive exPlanations (SHAP) method was introduced to explain the best models. ResultsOf the 885 patients enrolled, 221 (25.0%) developed POD. 22 (22.0%) of 100 validation cohort patients developed POD. The preoperative and postoperative artificial neural network (ANN) models exhibited optimal performance. The validation results demonstrated satisfactory predictive performance of the ANN model, with area under the receiver operator characteristic curve (AUROC) values of 0.776 and 0.684 for the preoperative and postoperative models, respectively. Based on the ANN algorithm, we constructed dynamic, highly accurate, and interpretable web risk calculators for POD. ConclusionsWe successfully developed online interpretable dynamic ANN models as clinical decision aids to identify patients at high risk of POD before and after cardiac surgery to facilitate early intervention or care.

Dynamic artificial neural network model for ultralow temperature prediction in hydrogen storage tank

Hydrogen fuel cell vehicles are gaining popularity as an alternative to electric vehicles. To support their use, research on hydrogen-storage devices and charging stations is being conducted actively. However, high-pressure hydrogen dispensing can lead to a temperature increase and potential explosion hazards. Therefore, maintaining safe hydrogen temperatures during charging is crucial. However, accurately predicting temperature changes in hydrogen heat exchangers is challenging owing to rapidly changing operating conditions and large temperature differences. In this study, we adopt a deep-learning-based artificial neural network for predicting temporal temperature variations in low-temperature hydrogen heat exchangers. Two networks, namely, the deep neural network (DNN) and recurrent neural network (RNN), are constructed with Bayesian optimization based on the experimentally measured data from an ultralow-temperature hydrogen heat exchanger to predict temperature. Both models showed superior prediction accuracies of temperature within 3.11 % of mean absolute error (MAE) and heat transfer rates of the heat exchanger within 2.53 % of MAE. In particular, DNN modeling is recommended as a universal method for predicting the temperature when previous data are known, and RNN modeling is recommended for predicting the temperature for successive actions in a compressed-hydrogen system.

Prediction of chemical composition concentration in an urban area by Artificial Neural Networks

An artificial neural network (ANN) is one of the computational methods that, through the learning process and using simple processors called neurons, tries to give a mapping between the input variables and the output variables. In this paper, six different recurrent dynamic ANN models are proposed to predict the air pollutant concentrations in city of Tehran, Iran. There is no need to know the details of the governing phenomena to develop the ANN models.The chemical composition as the air pollutants are consisting of NO2, SO2, CO, O3, PM10, and PM2.5. The proposed models are designed with an input layer consisting of meteorological variables and previous sampling times of each output variable. The models have the capability for air pollutant concentration prediction for 24 hours later. The results show that the developed models for NO2, SO2, CO, O3, PM10, and PM2.5 have the values of coefficient of determination (­R2) equal to 0.91, 0.95, 0.94, 0.97, 0.94, and 0.93. Also, the Normalized Root Mean Square Error (NRMSE) is equivalent to 0.0355, 0.2577, 0.047, 0.0397, 0.0270, and 0.0445 for NO2, SO2, CO, O3, PM10, and PM2.5 models, respectively.&#x0D; The results show that they all models have high accuracy and a low error value. The results may be because the model uses an almost complete set of input variables. Also the model used three sampling times as input variables through a recurrent structure to capture the dynamic behavior. The number of sampling times used as input variables through a recurrent structure that is related to the dynamic conditions of the model.

Time-delay neural network modeling of the running-in wear process

A time-delay neural network (TDNN) model is developed for describing the variation of the wear rate during the running-in wear process under constant relative sliding velocity. Besides the current value of the normal load, the neural network inputs also include its several preceding values. A specific feature of the TDNN model is that, by analogy with the Archard–Kragelsky wear equation, the neural network incorporates a power-law activation in the output layer for the current load input, which is directly fed into the output neuron along with the outputs of the hidden neurons. A motivation for the choice of such a dynamic artificial neural network model is given, and an intuitive explanation for the TDNN operation is suggested. A predictive capability of the TDNN model is illustrated based on experimental data taken from the literature.

Artificial Neural Network Travel Time Prediction Model for Buses Using Only GPS Data

Real-time and accurate travel time information of transit vehicles is valuable as it allows passengers to plan their trips to minimize waiting times. The objective of this research was to develop a dynamic artificial neural network (ANN) model that can provide accurate prediction of bus travel times to give real-time information at a given downstream bus stop using only global positioning system (GPS) data. The ANN model is trained off-line but can be used to provide real-time travel time information. To achieve this, care was taken in selecting a unique set of input-output combinations for prediction. The results obtained from the case study are promising to implement an Advanced Public Transportation System (APTS). The performance of the proposed ANN model was compared with a historical average model under two criteria: prediction accuracy and robustness. It was shown that the ANN outperformed the average approach in both aspects.

Dynamic state feedback controller and observer design for dynamic artificial neural network models

Artificial neural networks are black-box models that can be used to model nonlinear dynamical systems. This article presents a synthesis method for full dynamic state feedback controllers and state and output observers that have guaranteed properties for systems approximated by dynamic artificial neural networks. The resulting control designs are applicable to the practical situation in which the steady-state values for the control input are not known. Dynamic artificial neural networks are written in the standard nonlinear operator form, also known in the literature as the Luré formulation. A generalized form of the Luré formulation is adopted to allow for the representation of deep ℓ-layer networks, ℓ≥1. Sufficient conditions for controller synthesis and observer design are derived in the form of linear matrix inequalities, using a quadratic Lyapunov function. The synthesis method is demonstrated for the control of pH in two tanks in series and a numerical example.

Meta-ANN – A dynamic artificial neural network refined by meta-learning for Short-Term Load Forecasting

In this paper a dynamic Artificial Neural Network (ANN) model called Meta-ANN is developed for forecasting the short-term grid load. The primary ingredient of the model is a base module which is an ANN trained over a large historical data set to learn the long-term trend and seasonality of grid load. To capture the nonstationary pattern of the grid load, an error-correction module based on the idea of meta-learning is integrated into the model. This module finetunes the base module according to most recent prediction errors. For each day of interest, Meta-ANN generates a new ANN model started from the base module by tracing the gradient of the prediction loss on recent observations weighted by learning rates with specific structures. The full Meta-ANN model is trained by jointly optimizing the base module and error-correction module via gradient descent algorithms. The implementation based on gradient descent algorithms is detailed with streamlined mathematical formulations. The proposed model is tested on the open-access data from Elia, a Belgian transmission system operator, for forecasting the daily mean load and load profile. The numerical study shows that Meta-ANN makes more accurate and robust prediction by effectively capturing the nonstationary pattern in grid loads.

Forecasting stock returns on the Amman Stock Exchange: Do neural networks outperform linear regressions?

Different models have been used in the finance literature to predict the stock market returns. However, it remains an open question whether non-linear models can outperform linear models while providing accurate predictions for future returns. This study examines the prediction of the non-linear artificial neural network (ANN) models against the baseline linear regression models. This study aims specifically to compare the prediction performance of regression models with different specifications and static and dynamic ANN models. Thus, the analysis was conducted on a growing market, namely the Amman Stock Exchange. The results show that the trading volume and interest rates on loans tend to explain the monthly returns the most, compared to other predictors in the regressions. Moreover, incorporating more variables is not found to help in explaining the fluctuations in the stock market returns. More importantly, using the root mean square error (RMSE), as well as the mean absolute error statistical measures, the static ANN becomes the most preferred model for forecasting. The associated forecasting errors from these metrics become equal to 0.0021 and 0.0005, respectively. Lastly, the analysis conducted with the dynamic ANN model produced the highest RMSE value of 0.0067 since November 2018 following the amendment to the Jordanian income tax law. The same observation is also seen since the emerging of the COVID-19 outbreak (RMSE = 0.0042).

Prediction of time-varying inner wall temperature of surge lines by a dynamic neural network

Thermal fatigue, mainly caused by thermal stratification, is one of the major factors for reliability degradation of surge lines in nuclear power plants and is well managed through on-line temperature monitoring. However, the vast amount of outer wall temperature data generated by on-line temperature monitoring requires an efficient way of calculating the inner wall temperature. In this paper, the time-varying inner wall temperature is predicted rapidly and precisely using a dynamic artificial neural network model based on outer wall temperature and flow velocity. By optimizing the length of time series, the time delay of heat transfer between the inner wall and outer wall temperature is fully considered in the dynamic model. Consequently, the best root mean squared error of the temperature prediction is as low as 3.78. More importantly, the importance analysis based on dynamic model is also performed to quantify the importance of outer wall temperature variables, which can be used to optimize the arrangement of outer wall temperature measurement points for further applications under different circumstances to reduce cost while maintain high prediction accuracy. The results in this study will be helpful for the optimization of measurement points arrangement and rapid on-line thermal fatigue evaluation of nuclear power plants.

Soft Computing of a Medically Important Arthropod Vector with Autoregressive Recurrent and Focused Time Delay Artificial Neural Networks.

Simple SummaryArthropod vectors are responsible for transmitting a large number of diseases, and for most, there are still not available effective vaccines. Vector disease control is mostly achieved by a sustained prediction of vector populations to maintain support for surveillance and control activities. Mathematical models may assist in predicting arthropod population dynamics. However, arthropod dynamics, and mosquitoes particularly, due their complex life cycle, often exhibit an abrupt and non-linear occurrence. Therefore, there is a growing interest in describing mosquito population dynamics using new methodologies. In this work, we made an effort to gain insights into the non-linear population dynamics of Culex sp. adults, aiming to introduce straightforward soft-computing techniques based on artificial neural networks (ANNs). We propose two kind of models, one autoregressive, handling temperature as an exogenous driver and population as an endogenous one, and a second based only on the exogenous factor. To the best of our knowledge, this is the first study using recurrent neural networks and the most influential environmental variable for prediction of the WNv vector Culex sp. population dynamics, providing a new framework to be used in arthropod decision-support systems.A central issue of public health strategies is the availability of decision tools to be used in the preventive management of the transmission cycle of vector-borne diseases. In this work, we present, for the first time, a soft system computing modeling approach using two dynamic artificial neural network (ANNs) models to describe and predict the non-linear incidence and time evolution of a medically important mosquito species, Culex sp., in Northern Greece. The first model is an exogenous non-linear autoregressive recurrent neural network (NARX), which is designed to take as inputs the temperature as an exogenous variable and mosquito abundance as endogenous variable. The second model is a focused time-delay neural network (FTD), which takes into account only the temperature variable as input to provide forecasts of the mosquito abundance as the target variable. Both models behaved well considering the non-linear nature of the adult mosquito abundance data. Although, the NARX model predicted slightly better (R = 0.623) compared to the FTD model (R = 0.534), the advantage of the FTD over the NARX neural network model is that it can be applied in the case where past values of the population system, here mosquito abundance, are not available for their forecasting.

Removal of Boron by a Modified Resin in Fixed Bed Column: Breakthrough Curve Analysis Using Dynamic Adsorption Models and Artificial Neural Network Model

Removal of Boron by a Modified Resin in Fixed Bed Column: Breakthrough Curve Analysis Using Dynamic Adsorption Models and Artificial Neural Network Model

Standard representation and unified stability analysis for dynamic artificial neural network models

An overview is provided of dynamic artificial neural network models (DANNs) for nonlinear dynamical system identification and control problems, and convex stability conditions are proposed that are less conservative than past results. The three most popular classes of dynamic artificial neural network models are described, with their mathematical representations and architectures followed by transformations based on their block diagrams that are convenient for stability and performance analyses. Classes of nonlinear dynamical systems that are universally approximated by such models are characterized, which include rigorous upper bounds on the approximation errors. A unified framework and linear matrix inequality-based stability conditions are described for different classes of dynamic artificial neural network models that take additional information into account such as local slope restrictions and whether the nonlinearities within the DANNs are odd. A theoretical example shows reduced conservatism obtained by the conditions.

Comparative Study of Static and Dynamic Artificial Neural Network Models in Forecasting of Tehran Stock Exchange

During the recent decades, neural network models have been focused upon by researchers due to their more real performance and on this basis, different types of these models have been used in forecasting. Now, there is a question that which kind of these models has more explanatory power in forecasting the future processes of the stock. In line with this, the present paper made a comparison between static and dynamic neural network models in forecasting (uninvariable) the return of Tehran Stock Exchange (TSE) index in order to find the best model to be used for forecasting this series. The data were collected daily from 26/11/2009 to 17/10/2014. The models examined in this study included two static models (Adaptive Neuro-Fuzzy Inference Systems and Multi-layer Feed-forward Neural Network MFNN) and a dynamic model (nonlinear neural network autoregressive model NNAR). The findings showed that based on the Mean Square Error and Root Mean Square Error criteria, ANFIS model had a much higher forecasting ability compared to other models.

Development of a dynamic artificial neural network model of an absorption chiller and its experimental validation

The aim of this paper is to present a methodology to model and evaluate the energy performance and outlet temperatures of absorption chillers so that users can have reliable information on the long-term performance of their systems in the desired boundary conditions before the product is installed. Absorption chillers' behaviour could be very complex and unpredictable, especially when the boundary conditions are variable. The system dynamic must therefore be included in the model. Artificial neural networks (ANNs) have proved to be suitable for handling such complex problems, particularly when the physical phenomena inside the system are difficult to model. Reliable “black box” ANN modelling is able to identify the system's global model without any advanced knowledge of its internal operating principles. Knowledge of the system's global inputs and outputs is sufficient. The methodology proposed was applied to evaluate a commercial absorption chiller. Predictions of the ANN model developed were compared, with a satisfactory degree of precision, to 2 days of experimental measures. These days were chosen to be representative of the real dynamic operating conditions of an absorption chiller. The neural model predictions are very satisfactory: absolute relative errors of the transferred energy are within 0.1–6.6%.

A dynamic artificial neural network for assessment of land-use change impact on warning lead-time of flood

Floods always require innovative models for flood forecasting. This paper proposes a dynamic artificial neural network (DANN) model for evaluating land-use change impact (LUCI) scenarios on weighted average of warning lead-time of flood (WAWLTF) in an urbanised watershed. The simulated floods of a calibrated HEC-HMS hydrological model were used for training and testing of DANN model. The features of proposed DANN's structure were determined by minimisation of a new flood forecasting error (FFE) index. Results showed that the proposed procedure was able to optimise features of DANN structure by minimising FFE and produced an appropriate DANN model for assessment of LUCI on WAWLTF. The results also denoted that practicing suitable watershed management in future may improve WAWLTF encouragingly but never compensates negative impact of urbanisation completely. In conclusion, the model can be used as an efficient tool in similar urbanised watershed for assessment of LUCI on WAWLTF.

Nonlinear Autoregressive Exogenous modeling of a large anaerobic digester producing biogas from cattle waste

In waste-to-energy plants, there is every likelihood of variations in the quantity and characteristics of the feed. Although intermediate storage tanks are used, but many times these are of inadequate capacity to dampen the variations. In such situations an anaerobic digester treating waste slurry operates under dynamic conditions. In this work a special type of dynamic Artificial Neural Network model, called Nonlinear Autoregressive Exogenous model, is used to model the dynamics of anaerobic digesters by using about one year data collected on the operating digesters. The developed model consists of two hidden layers each having 10 neurons, and uses 18days delay. There are five neurons in input layer and one neuron in output layer for a day. Model predictions of biogas production rate are close to plant performance within ±8% deviation.

Artificial Neural Network Modeling of Ball Mill Grinding Process

Grinding consumes around 2% of the energy produced in the world but existing methods of milling are very inefficient and use only 5% of the input energy for real size reduction rest is consumed by machine itself. Chrome ores are comminute, filtered, pelletized and sintered to use into submerged arc furnace for ferrochrome production. Variation in ore properties affects the particle size distribution during milling. Artificial neural network based model is developed to predict the particle size distribution of ball mill product using grinding data available for difference in grindability of Sukinda chromite ores. Input variables for model were ball size, ball load, ball-ore ratio, grinding time. Output was particle size distribution (+75 μm, -75 μm, +38 μm; -38 μm). Three different kinds of mathematical models have been compared to predict the particle size distribution. Finally a neural network based model was found most accurate. Dynamic artificial neural network model does not require any material constant and optimizes the mathematical correlation with better accuracy in a dynamic process. This methodology can be used to develop an online system to predict the ball mill performance to improve the performance of grinding circuit in mineral, metal and cement industry.

Comparison of the ARMA, ARIMA, and the autoregressive artificial neural network models in forecasting the monthly inflow of Dez dam reservoir

Summary The goal of the present research is forecasting the inflow of Dez dam reservoir by using Auto Regressive Moving Average (ARMA) and Auto Regressive Integrated Moving Average (ARIMA) models while increasing the number of parameters in order to increase the forecast accuracy to four parameters and comparing them with the static and dynamic artificial neural networks. In this research, monthly discharges from 1960 to 2007 were used. The statistics related to first 42 years were used to train the models and the 5 past years were used to forecast. In ARMA and ARIMA models, the polynomial was derived respectively with four and six parameters to forecast the inflow. In the artificial neural network, the radial and sigmoid activity functions were used with several different neurons in the hidden layers. By comparing root mean square error (RMSE) and mean bias error (MBE), dynamic artificial neural network model with sigmoid activity function and 17 neurons in the hidden layer was chosen as the best model for forecasting inflow of the Dez dam reservoir. Inflow of the dam reservoir in the 12 past months shows that ARIMA model had a less error compared with the ARMA model. Static and Dynamic autoregressive artificial neural networks with activity sigmoid function can forecast the inflow to the dam reservoirs from the past 60 months.