Clustering assisted artificial neural network for handling noisy big data: An application for estimation of parameter in combined mode conduction and radiation heat transfer

A pattern net assisted mapping artificial neural network (PAMANN) model for estimation of parameters in problem with large data (1300 × 121 matrix size) is reported. A pattern net-based multilayer perceptron neural network (MLPNN) model for clustering the data, followed by mapping MLPNN model for mapping the target with the input, is developed as PAMANN model. A heat transfer problem with combined mode conduction and radiation in porous medium is solved numerically, and is called direct model. In the inverse model, a PAMANN model is developed by using data generated through the direct model. The PAMANN model is able to estimate two parameters (extinction coefficient β and convective coupling P₂) after taking temperature profile as input. The model is tested for different number of neurons in hidden layer, and different levels of noise in input data. Twelve different algorithms are explored in training of mapping MLPNN, and compared for performance. Levenberg–Marquardt algorithm is found to estimate the parameters with high accuracy, but took high CPU time. Bayesian regularization is found to consume very high CPU time with moderate accuracy in estimation of parameters. Variations in hidden layer neuron number and noise in input data, were done to analyze the performance of mapping MLPNN with different training algorithms. Algorithms O-Step Secant, conjugate gradient with Polak-Ribiére updates, and conjugate gradient with Fletcher-Reeves updates are able to handle all variations of noise and number of neurons in hidden layer, with good accuracy of estimation and low CPU time consumption. Under high computational resource LM algorithm can be used for all cases. Up to 0.99132 value of regression coefficient is obtained in mapping MLPNN model with 15 hidden neurons, indicating the high accuracy of the model. With the help of PAMANN model, highly accurate (absolute error 1.78%) estimation of parameters is obtained. The model can handle upto 1% noise in input data, while giving accurate results.     

Genetic algorithm-assisted artificial neural network for retrieval of a parameter in a third grade fluid flow through two parallel and heated plates

Genetic algorithm (GA) has been used to determine important attributes of artificial neural network (ANN), such as number of neurons in different hidden layers and division of data for training, validation, and testing. The GA-assisted ANN (GAAANN) model was used to retrieve third grade fluid (TGF) parameter (A) in a TGF flow problem. The TGF was allowed to flow through two parallel plates, which were subjected to uniform heat flux. The least square method (LSM) was used to solve the governing equations, for specified boundary conditions. In this way, temperature profiles for different values of A were computed by LSM, constituting the direct part of the problem. In the inverse part, the GAAANN model was fed with a temperature profile as input and the corresponding value of A was obtained as output. Four different GAAANN model were developed, and a detailed analysis was done in retrieving the value of A by different GAAANN models. Two very important and commonly used algorithms: Levenberg-Marquardt (LM) and scaled conjugate gradient are explored for training of the neurons. The entire four GAAANN model were able to retrieve the value of A with different levels of accuracy.     

Numerical investigation of mixing length on performance of lobed forced mixer nozzles

Geometric models of a lobed mixer nozzle with variation of mixing length are created and corresponding flow fields are simulated using a steady Reynolds Averaged Navier–Stokes (RANS) equation with realizable k–? turbulence model. The numerical simulation results show that the mixing length of the nozzle has a relatively great influence on the development of streamwise vortices and the effective range for streamwise vortices to intensify mixing is about 0.5D distance from the lobe trailing edge. The change of mixing length has little effect on the thermal mixing efficiency. As for the total pressure recovery coefficient, the shorter the mixing length, the sharper the total pressure recovery coefficient curve. On the nozzle exit section, the thermal mixing efficiency increases at first and then slightly declines, and the total pressure recovery efficiency changes little as the mixing length increases. In addition, the thrust coefficient has some small relationship with the mixing length. The thrust coefficient increases in waves as the mixing length increases and the difference between the maximum and the minimum values is slightly less than 1% when the mixing length increases from 0.4D to D. © 2011 Wiley Periodicals, Inc. Heat Trans Asian Res; Published online in Wiley Online Library wileyonlinelibrary.com/journal/htj . DOI 10.1002/htj.20343     

Daily means ambient temperature prediction using artificial neural network method: A case study of Turkey

The objective of this paper is to develop an artificial neural network (ANN) model which can be used to predict daily mean ambient temperatures in Denizli, south-western Turkey. In order to train the model, temperature values, measured by The Turkish State Meteorological Service over three years (2003–2005) were used as training data and the values of 2006 were used as testing data.In order to determine the optimal network architecture, various network architectures were designed; different training algorithms were used; the number of neuron and hidden layer and transfer functions in the hidden layer/output layer were changed. The predictions were performed by taking different number of hidden layer neurons between 3 and 30. The best result was obtained when the number of the neurons is 6. The selected ANN model of a multi-layer consists of 3 inputs, 6 hidden neurons and 1 output. Training of the network was performed by using Levenberg–Marquardt (LM) feed-forward backpropagation algorithms. A computer program was performed under Matlab 6.5 software. For each network, fraction of variance (R²) and root-mean squared error (RMSE) values were calculated and compared. The results show that the ANN approach is a reliable model for ambient temperature prediction.     

Numerical study on the SNCR application of space-limited industrial boiler

Considering that the direct application of SNCR method to existing industrial boiler is usually known as not quite satisfactory mainly due to the insufficient residence time by the limitation of SNCR application space, it is quite interesting to examine whether it is possible to improve the efficiency of SNCR reaction by the adjustment of design and operational parameters. Especially one of novel concepts is the introduction of auxiliary air employed for the increase of turbulence mixing of reduction material in SNCR injection system. To this end a comprehensive computer program is developed using SIMPLE algorithm by Patankar in order to evaluate the efficiency of SNCR system for the boiler with 40 tons of steam/h using heavy oil.The computer program is made in 3-D rectangular coordinate using various phenomenological models. For example, standard k–ε turbulence model and typical eddy breakup model are incorporated for the Reynolds stresses and turbulent reaction of major fuel species, respectively. However, detailed data of chemical kinetics are also incorporated for the process of NO formation together with the NH₃ reduction reaction of SNCR system. The complex coupling phenomena between combustion chemistry and turbulence are resolved by the strategy of harmonic mean expression assuming proper empiricism. Further, the calculation of droplet trajectory and volatilization is incorporated in Lagrangian frame assuming eight possible trajectories in each grid by the consideration of the calculation efficiency and accuracy.The validation of program developed was made by the comparison with the measured temperature profile and a series of parametric investigations have been performed to enhance the NO removal efficiency. The major variables considered in this study are droplet diameter, the injection location and amount of reduction agent together with the introduction of auxiliary mixing-enhanced air for the reduction material. Based on the results of the calculation, it is found that the removal efficiency of NO was improved to a significant amount by the increase of penetration depth of the reducing agent into the center region of boiler. The increase of penetration distance and thereby the enhancement of mixing efficiency was obtained either by the employment of the mixing air together with the increase of the injection velocity and droplet size, respectively. Based on this study, the successful application of SNCR method even for the space-limited industrial boiler can be achieved by the proper increase of the mixing of reducing agents.     

Performance prediction of wet cooling tower using artificial neural network under cross-wind conditions

This paper describes an application of artificial neural networks (ANNs) to predict the thermal performance of a cooling tower under cross-wind conditions. A lab experiment on natural draft counter-flow wet cooling tower is conducted on one model tower in order to gather enough data for training and prediction. The output parameters with high correlation are measured when the cross-wind velocity, circulating water flow rate and inlet water temperature are changed, respectively. The three-layer back propagation (BP) network model which has one hidden layer is developed, and the node number in the input layer, hidden layer and output layer are 5, 6 and 3, respectively. The model adopts the improved BP algorithm, that is, the gradient descent method with momentum. This ANN model demonstrated a good statistical performance with the correlation coefficient in the range of 0.993–0.999, and the mean square error (MSE) values for the ANN training and predictions were very low relative to the experimental range. So this ANN model can be used to predict the thermal performance of cooling tower under cross-wind conditions, then providing the theoretical basis on the research of heat and mass transfer inside cooling tower under cross-wind conditions.     

Modeling of a convective-infrared kiwifruit drying process

This paper aims to evaluate the experimental performance of a convective-infrared system with heat recovery (CIRHR) at different drying temperatures (40, 45, 50 and 55 °C) and 0.5 m/s air velocity and also to discuss and predict the performance of system on energy consumption and drying kinetics of sliced kiwifruit using artificial neural networks (ANNs). The energy efficiency values were obtained between 2.85% and 32.17%. The ANN model was used to predict the energy consumption of the system and moisture content of the kiwifruit. The back-propagation learning algorithm with Levenberg–Marquardt (LM) and Fermi transfer function were used in the network. The coefficient of determination (R²), the root means square error (RMSE) and the mean absolute percentage error (MAPE) were calculated as 0.99, 0.001 and 0.34, respectively. It can be concluded that predicted values are in good agreement with experimental results.     

Convective heat transfer and pressure drop study on nanofluids in double-walled reactor by developing an optimal multilayer perceptron artificial neural network

The effect of nanofluid on the cooling performance and pressure drop of a jacked reactor has experimentally been investigated. Aqueous nanofluids of Al₂O₃ and CuO was used as the cool ant inside the cooling jacket of the reactor. The application of the artificial neural networks (ANNs) to predict the performance of a double-walled reactor has been studied. Different architectures of artificial neural networks were developed to predict the convective heat transfer and pressure drop of nanofluids. The experimental results are used for training and testing the ANNs based on two optimal models via feed-forward back-propagation multilayer perceptron (MLP). The comparison of statistical criteria of different network shows that the optimal structure for predicting the convective heat transfer coefficient is the MLP network with one hidden layer and 10 neurons, which has been trained with Levenberg–Marquardt (LM) algorithm. The predicted pressure drop values by the MLP network with two hidden layers and 6 neurons in the each layer has been used from LM training algorithm, which showed a reasonable agreement with the experimental results.     

Artificial neural networks for automotive air-conditioning systems performance prediction

In this study, ANN model for a standard air-conditioning system for a passenger car was developed to predict the cooling capacity, compressor power input and the coefficient of performance (COP) of the automotive air-conditioning (AAC) system. This paper describes the development of an experimental rig for generating the required data. The experimental rig was operated at steady-state conditions while varying the compressor speed, air temperature at evaporator inlet, air temperature at condenser inlet and air velocity at evaporator inlet. Using these data, the network using Lavenberg–Marquardt (LM) variant was optimized for 4–3–3 (neurons in input–hidden–output layers) configuration. The developed ANN model for the AAC system shows good performance with an error index in the range of 0.65–1.65%, mean square error (MSE) between 1.09 × 10^?5 and 9.05 × 10^?5 and the root mean square error (RMSE) in the range of 0.33–0.95%. Moreover, the correlation which relates the predicted outputs of the ANN model to the experimental results has a high coefficient in predicting the AAC system performance.     

Predicting the accuracy of nanofluid heat transfer coefficient's computational fluid dynamics simulations using neural networks

This research presents a neural network algorithm to identify the best modeling and simulation methods and assumptions for the most widespread nanofluid combinations. The neural network algorithm is trained using data from earlier nanofluid experiments. A multilayer perceptron with one hidden layer was employed in the investigation. The neural network algorithm and data set were created using the Python Keras module to forecast the average percentage error in the heat transfer coefficient of nanofluid models. Integer encoding was used to encode category variables. A total of 200 trials of different neural networks were taken into consideration. The worst-case error bound for the chosen architecture was then calculated after 100 runs. Among the eight models examined were the single-phase, discrete-phase, Eulerian, mixture, the mixed model of discrete and mixture phases, fluid volume, dispersion, and Buongiorno's model. We discover that a broad range of nanofluid configurations is accurately covered by the dispersion, Buongiorno, and discrete-phase models. They were accurate for particle sizes (10–100 nm), Reynolds numbers (100–15,000), and volume fractions (2%–3.5%). The accuracy of the algorithm was evaluated using the root mean square error (RMSE), mean absolute error (MAE), and R² performance metrics. The algorithm's R² value was 0.80, the MAE was 0.77, and the RMSE was 2.6.     

Numerical investigation into thermal mixing efficiency in Y-shaped channel using Lattice Boltzmann method and field synergy principle

This study employs the lattice Boltzmann method to simulate the thermal mixing efficiency of two-dimensional, incompressible, steady-state low Reynolds number flows in a Y-shaped channel. The effects of introducing a staggered arrangement of wave-like and circular obstacles into the mixing section of the Y-shaped channel are systematically examined. The simulation results demonstrate that both types of obstacle yield an effective improvement in the thermal mixing efficiency compared to that achieved in a Y-channel with a straight mixing section. Adopting the field synergy principle, it is demonstrated that the enhanced mixing efficiency is the result of an increased intersection angle between the velocity vector and the temperature gradient within the channel.     

基于GA-LMBP神经网络模型的大坝渗流压力预报分析

为克服传统BP神经网络在渗流压力预测过程中收敛慢、计算量大和易陷入局部极小等缺陷,依据渗流压力的影响因素,研究了模型的结构和输入输出因子,建立了基于遗传算法和LM算法相结合的GA-LMBP神经网络的大坝渗流压力预测模型,即通过遗传算法(GA)的选择、交叉和变异操作得到BP网络的一组全局最优近似解(即网络的初始权值和阈值),再以该近似解为初值,利用LM算法对BP网络进行优化训练,将训练好的网络用于渗流压力的预测。实例应用结果表明,在相同精度的要求下,GA-LMBP神经网络模型收敛速度快、预测精度高,对大坝渗流压力的预测效果更佳,是值得采用的一种模型。     

Predicting and optimizing the heat transfer performance of radiator with functionalized graphene nanofluids

Accurately predicting the heat transfer characteristics of coolants used in thermal management of energy systems like heat exchangers, power electronics, and heating, ventilation, and air conditioning is indispensable in maintaining its operating conditions within safety limits. Apart from safety, factors such as power consumption and operating cost are the most important constraints to be considered in designing an energy-efficient and cost-effective cooling solution. In this study, the experimental data available from previous research on the use of functionalized graphene-based nanofluids in compact heat exchangers such as the automotive radiator is used to optimize the heat transfer performance parameters like Nusselt number of the nanofluid, the friction factor, and effectiveness of the heat exchanger. A supervised machine learning technique like the artificial neural network is used to obtain the objective functions of the response variables in terms of input features such as Reynolds number, Prandtl number, the volume concentration of nanoparticles in the base fluid, number of transfer units, heat capacity, the density of nanofluid, pressure drop and velocity. On the current dataset, it is found that by using the Bayesian regularization training algorithm and tangent sigmoidal activation function in the neural network, the best accuracies in the prediction can be achieved. Well-known nature-inspired optimization algorithms like genetic algorithms and simulated annealing are used in optimizing the above-mentioned response variables. Both algorithms converged to the same values of the objective functions. The optimum values of Nusselt number, effectiveness, and friction factor are 105.65, 0.506, and 0.0038, respectively, for the given composition of the nanofluid and radiator configuration.     

Thermal performance of packed bed thermal energy storage units using multiple granular phase change composites

The present article reports on the utilization of multiple granular phase change composites (GPCC) with different ranges of phase change temperatures in a packed bed thermal energy storage system. Small particle diameter of GPCC allows simple mixing of two or three ranges of GPCCs in a packed bed for enhancement of storage unit performance. Experiments have been carried out to characterize the phase changing characteristics of two GPCCs chosen for this purpose. Packed bed column experiments have been carried out to provide basic understanding of the heat transfer process in the composite bed consisting of a mixture of GPCCs at different values of mixing ratio. A mathematical model has been developed for the analysis of charging and discharging process dynamics. Once validated, the model has been used to perform a parametric study to investigate the overall bed performance at different values of mixing ratio and Reynolds number. An optimization of the value of mixing ratio has been obtained based on the overall charging and discharging times as well as the exergy efficiency. It has been demonstrated that, as compared to the use of single GPCC, careful choice of the mixing ratio of GPCCs in a composite bed can result in a significant enhancement of the overall storage unit performance. As compared to the use of multiple sequential layers of GPCCs, using units composed of a mixture of GPCCs with an optimized mixing ratio results in a remarkable improvement of the unit performance without limitations on the charging and discharging directions during practical applications.     

主支管动量比对T型管道流体混合过程影响

采用大涡模拟(large-eddy simulation,LES)的方法对T型管道内主管与支管不同动量比的流体混合过程的流动情况进行了数值模拟,采用时均值和均方根值来描述速度的平均大小和波动强度。通过改变主管和支管的速度比即动量比,将流体分为三类:碰撞射流、偏射流和壁面射流,研究其对速度的平均值和波动的影响,并研究其所反映的惯性力对流动的影响。该研究揭示了流体混合过程中动量比对波动的影响规律,对预测和校核管壁疲劳失效具有重要的指导意义。     

A numerical study to investigate the effect of turbulators on thermal flow and heat performance of a 3D pipe

To reduce the heat exchanger's costs in a highly competitive industry, thermal performance enhancement of the heat exchangers has successfully gained attention in the last few decades. Among different engineering approaches, the application of the enhanced pipes provides a key solution to improve heat performance. In this paper, the investigation develops a numerical study based on the commercially available computational fluid dynamics codes on the turbulent flow in three-dimensional tubular pipes. Various concavity (dimple) diameters with corrugation and twisted tape configurations are investigated. The study has shown that perforated geometrical parameters lead to a high fluid mixing and flow perturbation between the pipe core region and the walls, hence better thermal efficiency. Moreover, a model of concavity (dimple) with a 4 mm diameter allows the highest heat transfer enhancement among other designs. In addition, the study shows that due to the disturbance between the pipe core region and the pipe wall, the transverse vortices and swirl flow generated are forceful, which leads to better heat transfer enhancement compared with the conventional (smooth) pipes. As the Reynolds number (Re) rises, the mixing flow, secondary, and separation flow extend to become higher than the values in a smooth pipe, allowing a higher value of performance evaluation factor to be achieved for a dimple diameter of 1mm at the low Re values. This study, therefore, shows the promising potential of the enhanced pipes in the heat transfer enhancement of heat exchangers that is crucial in industrial applications to save more energy.     

Parabolic RANS solver for low‐computational‐cost simulations of wind turbine wakes

A numerical framework for simulations of wake interactions associated with a wind turbine column is presented. A Reynolds‐averaged Navier‐Stokes (RANS) solver is developed for axisymmetric wake flows using parabolic and boundary‐layer approximations to reduce computational cost while capturing the essential wake physics. Turbulence effects on downstream evolution of the time‐averaged wake velocity field are taken into account through Boussinesq hypothesis and a mixing length model, which is only a function of the streamwise location. The calibration of the turbulence closure model is performed through wake turbulence statistics obtained from large‐eddy simulations of wind turbine wakes. This strategy ensures capturing the proper wake mixing level for a given incoming turbulence and turbine operating condition and, thus, accurately estimating the wake velocity field. The power capture from turbines is mimicked as a forcing in the RANS equations through the actuator disk model with rotation. The RANS simulations of the wake velocity field associated with an isolated 5‐MW NREL wind turbine operating with different tip speed ratios and turbulence intensity of the incoming wind agree well with the analogous velocity data obtained through high‐fidelity large‐eddy simulations. Furthermore, different cases of columns of wind turbines operating with different tip speed ratios and downstream spacing are also simulated with great accuracy. Therefore, the proposed RANS solver is a powerful tool for simulations of wind turbine wakes tailored for optimization problems, where a good trade‐off between accuracy and low‐computational cost is desirable.     

A dynamic two-level artificial neural network for estimation of parameters in combined mode conduction-radiation heat transfer in porous medium: An application to handle huge dataset with noise

A dynamic two-level artificial neural network (DTLANN) approach is used for the estimation of parameters in combined mode conduction–radiation heat transfer in a porous medium. Four commonly used neural networks: feed forward, cascade forward, fitnet, and radial basis are used in mapping artificial neural network (ANN), and their performance is compared under noisy big data (10,302 × 1300 matrix size). Governing equations for heat transfer in the porous medium through conduction and radiation modes are solved by finite volume method and discrete transfer method. This numerical model is called a direct model. A large amount of data is generated by using the direct model for different values of extinction coefficient β and convective coupling P₂. These data were divided into different groups (class) based on the temperature difference between the gas and solid phase. In the inverse analysis, a new pair of temperature profiles for the solid and gas phase is taken as input and classified with the help of a pattern net artificial neural network model. On the basis of classification, data from that particular class and its neighbor class are used for training the mapping ANN model. After the training of the mapping ANN model, corresponding values of β and P₂ are obtained as output for any new input. This DTLANN model has a high regression coefficient (R) of .99131 and can predict highly accurate values of parameters under a huge dataset with noise, within much less CPU time.     

Inverse neural network for optimal performance in polygeneration systems

In this paper, inverse neural network (ANNi) is applied to optimization of operating conditions or parameters in energy processes. The proposed method ANNi is a new tool which inverts the artificial neural network (ANN), and it uses a Nelder-Mead optimization method to find the optimum parameter value (or unknown parameter) for a given required condition in the process. In order to accomplish the target, first, it is necessary to build the artificial neural network (ANN) that simulates the output parameters for a polygeneration process. In general, this class of ANN model is constituted of a feedforward network with one hidden layer to simulate output layer, considering well-known input parameters of the process. Normally, a Levenberg–Marquardt learning algorithm, hyperbolic tangent sigmoid transfer-function, linear transfer-function and several neurons in the hidden layer (due to the complexity of the process) are considered in the constructed model. After that, ANN model is inverted. With a required output value and some input parameters it is possible to calculate the unknown input parameter using the Nelder-Mead algorithm. ANNi results on three different applications in energy processes showed that ANNi is in good agreement with target and calculated input data. Consequently, ANNi is applied to determine the optimal parameters, and it already has applications in different processes with a very short elapsed time (seconds). Therefore, this methodology can be useful for the controlling of engineering processes.     

一种基于MLP神经网络的农用井泵故障诊断模型及应用

针对农用井泵常见的故障,建立了基于前馈MLP神经网络的故障诊断模型,并用反向传播算法进行训练。同时,用隶属度值定量描述故障程度,以出水量变小、有震动或噪声、出浑水、水泵不转、密封处温度异常升高5种常见的异常状况和井龄、泵龄、扬程、水泵淹没深度4个辅助条件作为网络的输入,以12种故障原因的4位二进制码和程度值作为网络输出,通过公式法确定隐含层的节点个数为10。通过训练后的神经网络对测试样本进行检验,验证了该神经网络模型对于农用井泵故障诊断的适用性。