Modeling of an industrial drying process by artificial neural networks

Assidjo, Emmanuel; Yao, Benjamin; Kisselmina, K.; Amane, Dhanashree

doi:10.1590/s0104-66322008000300009

Cited by 21 publications

(11 citation statements)

References 21 publications

(17 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…According to Assidjo et al (2008) [41], the best network is a compromise between the results obtained during training and those with the generalization (test): the network that best characterizes the data of the phenomena is the network with the least error. In this study, the best compromise (providing both a low value of MSE and a high value of R for the training and test sets) was obtained with the 3-7-2 structure ANN model depicted in Figure 2.…”

Section: Global Modeling and Optimizationmentioning

confidence: 99%

Optimization of Rubber Seed Oil Transesterification to Biodiesel Using Experimental Designs and Artificial Neural Networks

Kouassi¹,

Abolle²,

Yao³

et al. 2018

GSC

View full text Add to dashboard Cite

The development of biofuels is driven both by concern about the greenhouse effect and by interest in the opportunities for exploitation of biomass of agricultural origin. In order to improve the yield and quality of biodiesel through modeling and optimization, several studies are in progress. In this paper, biodiesel produced from rubber seed oil in the homogeneous transesterification is studied using a Plackett-Burman experimental design, a full factorial design, a central composite design and an Artificial Neural Network (ANN) coupled with a Genetic Algorithm (GA).Variables such as temperature, stirring speed, reaction time, type of alcohol, and type of catalyst are studied to obtain the best specific gravity and kinematic viscosity. Type of alcohol and type of catalyst have the greatest effect on the two responses, with ethanol (alcohol) and sulphuric acid (catalyst) producing the best results. The specific gravity and kinematic viscosity changes recorded during the transesterification process followed the first and second order polynomial models, respectively. The ANN coupled with GA was used to optimize the two responses simultaneously. Global optimal values of specific gravity (0.883) and kinematic viscosity (6.76 cSt) were recorded when a temperature of 90˚C, a stirring speed of 305 rpm, and a treatment time of 141 min were imposed.

show abstract

Section: Global Modeling and Optimizationmentioning

confidence: 99%

Optimization of Rubber Seed Oil Transesterification to Biodiesel Using Experimental Designs and Artificial Neural Networks

Kouassi¹,

Abolle²,

Yao³

et al. 2018

GSC

View full text Add to dashboard Cite

show abstract

“…support vector machine (SVM), self-organisation map (SOM), multilayer perceptron (MLP)). The third (MLP) is the most widely used (Assidjo et al 2008). A multilayer feed-forward neural network, have three type layers that including input, hidden layer and output.…”

Section: B Artificial Neural Networkmentioning

confidence: 99%

“…Artificial neural networks (ANN) model is able to handle complex nonlinear relationships with ease, even when the exact nature of such relationships is unknown [7,8]. Artificial neural networks (ANN) used for modeling the drying step of the production by references [9,10,11]. This method can satisfactorily model the process and the predictions of the artificial neural networks model fit the experimental data more accurately in comparison to the various mathematical equations.…”

Section: Introductionmentioning

confidence: 99%

Modeling of moisture content in tomato drying procces by ANN-GA technique

Javadikia

Rafiee

Taheri‐Garavand

et al. 2011

2011 1st International eConference on Computer and Knowledge Engineering (ICCKE)

View full text Add to dashboard Cite

A Feed Forward Neural Network (FFNN) is designed to estimate the moisture content of dried tomato. The experiment is done by a dryer that it was capable of providing any desired drying air temperature, relative humidity and velocity. After getting the practical data, a general FFNN is designed and optimized with Genetic Algorithm (GA) through MATLAB software. Result showed that the configuration of FFNN and GA is very powerful and it is able to model any set of data. Finally, the result of best network by GA was a network with only one hidden layer and 11 neurons and this network could predict moisture content of dried tomato with correlation coefficient of 0.99.

show abstract

“…Since then, many aspects of drying processes have been modeled by this tool, such as drying kinetics [20][21][22] , moisture sorption isotherms [23] , product properties [24][25] etc. A review on the use of artificial neural networks in drying technology was recently published [26] .…”

Section: Introductionmentioning

confidence: 99%

Control of the Moisture Content of Milk Powder Produced in a Spouted Bed Dryer Using a Grey-Box Inferential Controller

Vieira

Freire

2015

Drying Technology

View full text Add to dashboard Cite

Although the moisture content of dried products is an important variable on industrial dryers, it is often not measured directly for control purposes. Alternative and simpler meters might provide information to be used by a physical-mathematical model to estimate the moisture content. When this procedure is applied to a control strategy, an inferential controller is developed. In this paper, a physical-mathematical model was used to infer the moisture content of milk powder produced in a spouted bed dryer.Afterwards, simulations of an inferential proportional-integral controller were carried out, using the inlet air heating rate as the manipulated variable. The physical-mathematical model used in the procedure was a hybrid one, which considers mass and energy balances and one term which is estimated by an artificial neural network. The controller parameters (controller gain and integral time) were tuned by trial and error. Even though the procedure was quite simple, it was proven to be effective in yielding a stable closedloop response for both servo and regulatory control of the (inferred) powder moisture content.

show abstract

Modeling of an industrial drying process by artificial neural networks

Cited by 21 publications

References 21 publications

Optimization of Rubber Seed Oil Transesterification to Biodiesel Using Experimental Designs and Artificial Neural Networks

Optimization of Rubber Seed Oil Transesterification to Biodiesel Using Experimental Designs and Artificial Neural Networks

Modeling of moisture content in tomato drying procces by ANN-GA technique

Control of the Moisture Content of Milk Powder Produced in a Spouted Bed Dryer Using a Grey-Box Inferential Controller

Contact Info

Product

Resources

About