Multilayer Feed Forward Model Research Articles

Research on temperature performance prediction of vortex tubes based on artificial neural networks

Abstract This study constructs a hybrid neural network model by integrating the physical constraints of the Bernoulli equation and Nikolaev's formula. The model is designed to explore and predict the variation pattern of the cold end temperature in a vortex tube. The input parameters include inlet pressure, inlet temperature, and cold mass fraction, with the cold end temperature as the output parameter. The network employs a multilayer feedforward model and the Levenberg-Marquardt learning algorithm, using a hyperbolic tangent function as the activation function. To evaluate the statistical validity of the developed model, the coefficient of determination (R²) and root mean square error (RMSE) are utilized, along with an analysis of the model's uncertainty and robustness. The hybrid model achieves an R² of 0.9936 and an RMSE of 0.3392, demonstrating strong performance in terms of uncertainty and robustness. These results indicate that the model accurately predicts the cold end temperature variation in the vortex tube. Furthermore, the findings reveal an optimal pressure range (0.49 MPa to 0.76 MPa) and cold mass fraction range (0.1 to 0.2) that minimize the cold end temperature.

Artificial Neural Network Model to Forecast Energy Consumption in Wheat Production in India

Energy analysis in agriculture sector require modelling technique that can incorporate complex unknown interactions and non-linearity in systems. In this study Artificial neural network technique is used to model and forecast input energy consumed in wheat production in India and is compared for predictive accuracy with linear models. Household data from 256 farmers revealed that the average input energy consumed in region is 29612.43 MJ/ha with urea (47%), diesel (31.5%) and electricity (9.8%) being three main contributors. Multi-layered feed forward model with 2 hidden layers with 8 and 15 neurons respectively and sigmoidal activation function in hidden layers and output layers under gradient descent training algorithm gave the best results. The R2 was 0.99 for training dataset and 0.973 for validation data set, while for MLR model it was 0.95 and 0.73 for respective datasets. The root mean squared error (RMSE) in ANN model was 4779.2 MJ/ha and 2008.96 MJ/ha for training and validation data, respectively. This prediction system could forecast input energy with error margin of ± 7889.83 MJ/ha on training dataset and ± 3298.47 MJ/ha on validation data under various combinations. Sensitivity analysis showed that urea, diesel, and electricity are the important factors in input energy forecasting.

Prediction of the effect of temperature on electric power in photovoltaic thermal systems based on natural zeolite plates

The present study aims to predict the effect of the panel temperature on the electrical power obtained by the photovoltaic thermal system (PVT) based on natural zeolite plates. It was carried out using the long short-term memory (LSTM) and multilayer feed forward (MLF) algorithms, which are popular regression-based deep learning methods. Models have been developed that can predict the effect of temperature on electrical power by using the regression data obtained from experimental measurements on the PVT system integrated with natural zeolite plates. In the study, a polycrystalline PVT panel was chosen, in the designed models, temperature, solar radiation, and photovoltaic panel temperature values were evaluated as input, and the amount of power to be produced as output. Mean absolute error, root mean square error (RMSE), mean absolute percentage error (MAPE), and R2 functions were used for the prediction results of LSTM and MLF models designed with determined optimum hyperparameters. MAPE and R2 values for long short-term memory (LSTM) and MLF algorithms were determined as 0.100 and 0.996, as well as 0.131 and 0.923, respectively. As can be understood from these values, the effect of the designed temperature on the PVT panel can be calculated with a very small margin of error with the LSTM algorithm.

Multilayer Feed Forward Models in Groundwater Level Forecasting Using Meteorological Data in Public Management

Managing the groundwater resources is very vital for human life. This research proposes a methodology for predicting the groundwater levels which can be very valuable in water resources management. This study investigates the application of multilayer feed forward network models for forecasting the groundwater values in the region of Montgomery country in Pennsylvania. Multiple training algorithms and network structures were investigated to develop the best model in order to forecast the groundwater levels. Several multilayer feed forward models were created in order to be tested for their performance by changing the network topology parameters so as to find the optimal prediction model. The forecasting models were developed by applying different structures regarding the number of the neurons in every hidden layer and the number of the hidden network layers. The final results have shown a very good forecasting accuracy of the predicted groundwater levels. This research can be very valuable in water resources and environmental management.

A computational fluid dynamics based artificial neural network model to predict solid particle erosion

Solid particle erosion plays a critical role in the design and reliability of equipment employed in the oil and gas industry. Significant erosion occurs due to solid particle loading, especially in applications involving sand production. Low particle loading in drilling fluids (<10%) is also a source of erosion inside downhole tools and in rig equipment at the surface. Accurate prediction of erosion rates can save money and lives by predicting failure accurately and helping to maintain the safety of the equipment. Empirical and mechanistic models to predict erosion were primarily developed based on observations of extensive experiments and field studies. Computational fluid dynamics (CFD) has emerged as an alternative tool to predict erosion in recent years. The ability to simulate multiphase flows in complex geometries using CFD makes it a valuable and less-expensive method to predict erosion flow loop experimentations and field trials. Various empirical relations have been established to predict erosion using CFD. These methods often predict erosion regions accurately, but typically are highly inaccurate in predicting an erosion rate. An order-of-magnitude error is observed in many cases.This study employs machine learning approach along with CFD-based methodology to develop robust erosion models. A generalized model is developed based on experiments conducted on 90-degree elbows of 1-in. diameter and made from Inconel 718, Nickel Alloy 825, 25% Cr, Nickel Alloy 925, and 13% Cr L-80 materials. The Baker Hughes erosion model developed in 2008 is studied as a baseline. Statistical analysis was performed on CFD output parameters to identify those that most affect erosion rates. A correlation analysis and non-parametric statistical analysis is performed resulting in the development of two new regression models based on turbulent kinetic energy, and surface shear stress was developed. A 25-% improvement is observed in the predictions of cumulative erosion rate error compared to baseline. An artificial neural network with multilayer feed-forward model with the back-propagation algorithm and Levenberg-Marquardt training was developed. This model, along with Bayesian regularization, reduced cumulative error to less than 10%, compared to more than 40% in the baseline Baker Hughes model.

Modelling of Multi Layer Feed forward Neural Networks to Determine the Compressive Strength of Marmara Region Aggregate's Concrete

We aim to estimate concrete compressive strength by using the physical properties of the aggregates that are the main components of concrete. For this aim, concrete samples were prepared with aggregates having different origins and characteristics obtained from different 10 locations of the Marmara region. The compressive strength's results obtained by changing the aggregates were compared by ensuring the other components forming the concrete remained constant. 330 separate experiments were conducted to determine the physical characteristics of the aggregates, and these characteristics were used as input data in a multi layer Feed Forward Network model. The compressive strength of 7 and 28 days of the concrete obtained from the experiments were used as output in to a Multi-layer Feed Forward model. The training and test results from the models coincided closely with experimental results; also, the results were compared to the estimations made with a linear regression method.

Comparison of Neural Network Models in the Estimation of the Performance of Solar Collectors

AbstractThree artificial neural network models [feedforward, Elman, and nonlinear autoregressive exogenous (NARX)] were used to find the performance of two flat-plate collectors operating under Jordanian climate. One collector used water as a working fluid, while the other used fuel oil as a working fluid. Previously obtained experimental data on the performance of solar collectors were used to train the neural network. Density of fluid, input temperature, output temperature, ambient temperature, and solar radiation were used as input parameters in the input layer of the network while the efficiency of the flat-plate solar collector was in the output layer. It was found that the artificial neural network technique may be used to estimate the efficiency of the flat-plate collector with excellent accuracy. The obtained results showed that the multilayer feedforward model with five inputs has the best ability to estimate the required performance, while the other models, the feedforward, NARX, and Elman netwo...

Vehicle Positioning Using GSM and Cascade-Connected ANN Structures

Procuring location information for intelligent transportation systems is a popular topic among researchers. This paper investigates the vehicle location algorithm based on the received signal strength (RSS) from available Global System for Mobile Communications (GSM) networks. The performances of positioning models, which consisted of cascade-connected (C-C) artificial neural network (ANN) multilayer feedforward structures employing the space-partitioning principle, are compared with the single-ANN multilayer feedforward model in terms of accuracy, the number of subspaces, and other positioning relevant parameters. C-C ANN structures make use of the fact that a vehicle can be found only in a subspace of the entire environment (roads) to improve the positioning accuracy. The best-performing C-C ANN structure achieved an average error of 26 m and a median error of less than 5 m, which is accurate enough for most of the vehicle location services. Using the same RSS database obtained by measurements, it was shown that the proposed model outperforms <i xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">k</i> NN and extended Kalman filter (EKF)-trained ANN positioning algorithms. Moreover, the presented ANN structures replace not only the positioning algorithms but the overloaded map-matching process as well.

Using artificial neural network for predicting performance of the Ranque–Hilsch vortex tube

In this study, effects of conical valve angle and length to diameter ratio on the performance of a counter flow Ranque–Hilsch vortex tube are predicted with artificial neural networks (ANNs) by using experimental data. In the model, inlet pressure (Pi), conical valve angle (ϕ), length to diameter ratio (L/D) and cold mass fraction (yc) are used as input parameters while total temperature difference (ΔT) is chosen as the output parameter. The multilayer feed forward model and the Levenberg–Marquardt learning algorithm are used in the network and the hyperbolic tangent function is chosen as a transfer function. The artificial neural network is designed via the NeuroSolutions 6.0 software. Finally, it’s disclosed that ANN can be successfully used to predict effects of geometrical parameters on the performance of the Ranque–Hilsch vortex tube with a good accuracy.

Forecasting Influent-Effluent Wastewater Treatment Plant Using Time Series Analysis and Artificial Neural Network Techniques

We consider the problem of predicting the future behavior of wastewater treatment plant quality indicators by creating prediction models using historical plant data. One of the main aims of this work is to be able to predict plant operational situations in advance so that corrective actions can be taken in time. Sets of historical plant data, such as BOD, COD and TSS were collected for a local wastewater treatment plant in Doha, the capital of the State of Qatar. These variables characterize the performance of any wastewater treatment plant and can be considered as quality indicators of the plant performance. Data were collected over a period of 4 years for the influent and effluent streams of the station. The plant influent and effluent predictions were performed using different techniques. These include time-series analysis, where the ARIMA (Autoregressive Integrated Moving Average) model was implemented in this case, and two Artificial Neural Networks (ANN) algorithms, namely Adaptive Linear Neuron networks (ADALINE) and Multi-layer Feedforward (ML-FF) neural networks. The predictions from the three techniques were presented and compared. The ML-FF model predictions proved to be more reliable than that of the equivalent ARIMA predictions followed by the ADALINE predictions, particularly for the finial effluent stream variables.

Modelling hook times of mobile cranes using artificial neural networks

The hook times of mobile cranes are processes that are of non‐linear and discrete nature. Artificial neural network is a data processing technique that lends itself to this kind of problem. Three common artificial neural network architectures – multi‐layer feed‐forward (MLFF), group method of data handling (GMDH) and general regression neural network (GRNN) – are compared. The results show that the GRNN model aided with genetic algorithm (GA) is most promising in describing the non‐linear and discrete nature of the hook times. The MLFF model can also give a moderate level of accuracy in the estimation of hook travelling times of mobile cranes and is ranked second. The GMDH model is outperformed by the former two due to a less promising R‐square.