Original papers
Development of artificial intelligence based systems for prediction of hydration characteristics of wheat

https://doi.org/10.1016/j.compag.2016.08.014Get rights and content

Highlights

  • Three ANFIS models were developed to simulate hydration characteristics of wheat.

  • The results of ANFIS models were compared with those of ANN model.

  • Both ANFIS and ANN were able to accurately predict hydration characteristics.

  • An ANN simulation framework was preferred to three ANFIS simulation frameworks.

Abstract

Hydration characteristics (moisture content, moisture ratio and hydration rate) of wheat kernel during soaking process were studied and simulated. Hydration procedure was performed at five different experimental water temperatures (30, 40, 50, 60 and 70 (°C)) with respect to hydration time changes. Hydration characteristics of samples were modeled using one of the most popular simulation frameworks of artificial intelligence, adaptive neuro-fuzzy inference system (ANFIS). A comparison was also made between results of the best developed ANFIS model and those of the another well-known artificial intelligence technique, artificial neural network (ANN). The hydration temperature and time were used as input parameters and moisture content, moisture ratio and hydration rate were taken as output parameters of the intelligent modeling frameworks. To attain the best model with the highest predictive ability, developed models were compared based on statistical parameters of coefficient of determination, root mean square error and mean relative deviation modulus. According to the results, although the best framework of both ANFIS and ANN were able to accurately predict hydration characteristics, the best ANN simulation framework with a simple structure (2–4–3) was easier to use than the ANFIS with three different structures. ANN surface plots also illustrated that increasing the hydration temperature and time increased moisture content and decreased moisture ratio. Moreover, ANN modeling results indicated that the hydration rate increased in the initial and decreased in the middle and final phase of hydration procedure.

Introduction

Hydration is widely used as a pre-treatment for legumes and cereals before some processing operations such as extraction, cooking, germination, tempering and wet milling. It is necessary to know water absorption behavior in legumes and cereals during hydration because it affects following processing operations and the quality of the final product (Turhan et al., 2002).

In this context, there are a large number of researches reporting that many authors have investigated the hydration characteristics of different cereals and legumes at various conditions. Some of these attempts can be cited for rice (Kashaninejad et al., 2007), bean (Piergiovanni, 2011), corn kernels (Marques et al., 2014), lentil (Oroian, 2015) and soybean (Fracasso et al., 2015).

According to the literature, it can be inferred that the better understanding of the hydration kinetics of food products is achieved by modeling the hydration procedure accurately. Realizing the importance of this point resulted in development of several mathematical models to describe the hydration characteristics.

Peleg model is one of the well-known models frequently used by researchers to predict moisture content and hydration rate of agricultural products in hydration procedure (Maharaj and Sankat, 2000, Shittu et al., 2004, Cunningham et al., 2007, Kashiri et al., 2010, Shafaei et al., 2016). Also, Weibull model has commonly been applied in previous studies to predict moisture ratio of agricultural products (Marabi et al., 2003, Garcia-Pascual et al., 2006, Goula and Adamopoulos, 2009, Diaz-Ramirez et al., 2013, Rafiq et al., 2015). Other mathematical models for prediction of hydration characteristics of agricultural products can be found in the literature (Ibarz et al., 2004, Saguy et al., 2005, Wood and Harden, 2006, Vega-Galvez et al., 2009, Shafaei and Masoumi, 2014a, Shafaei and Masoumi, 2014b, Shafaei and Masoumi, 2014c, Oli et al., 2014, Ibarz and Augusto, 2015).

The hydration modeling has been performed in respect of hydration time using mathematical models for certain hydration temperature. The fitted models predicted the moisture content or moisture ratio based on one variable (hydration time). Coefficients of the models also changed with hydration temperature change. However, a comprehensive multiple variable model should be used for accurate prediction of hydration characteristics (moisture content, moisture ratio and hydration rate) of agricultural products during soaking based on multiple inputs (hydration time and temperature).

Development of intelligent models for prediction of multiple variable phenomena has resulted in general use of them in scientific and engineering practices. Unlike previously common modeling techniques, such as regression equations, artificial intelligence modeling techniques are applied to predict the target variable using more than one input and output. Moreover, such models are capable of determining nonlinear relationships existing between output and input variables. ANNs are data simulation frameworks based on the structure of the biological neural system. These intelligent models are mostly useful where old techniques fail to work correctly. ANFIS is also a new simulator tool used to increase the prediction efficiency of nonlinear relationships according to the theory of training algorithm and expert knowledge in a controlled environment.

Artificial intelligence modeling techniques have already been applied to simulate different processes. The literature includes studies which present the use of ANN to predict the nonlinear relationships in agricultural engineering applications (Adhikari and Jindal, 2000, Diamantopoulou, 2005, Lertworasirikul and Saetan, 2010, Torrecilla et al., 2016). In addition, a careful review of the literature shows the capability of ANFIS simulation environment in proper estimation of desired agricultural parameters (Brown-Brandl et al., 2005, Ghoush et al., 2008, Amiryousefi et al., 2011, Sefeedpari et al., 2014, Shafaei et al., 2015).

Some researchers have studied the hydration characteristics of wheat kernel from different points of view. Thus, several hydration conditions and modeling techniques were examined to optimize the performance of the related processing equipment (Bloome et al., 1982, Kang and Delwiche, 1999, Kang and Delwiche, 2000, Maskan, 2001, Maskan, 2002, Tagawa et al., 2003, Muramatsu et al., 2006, Kashaninejad and Kashiri, 2008, Kashaninejad et al., 2009, Vengaiah et al., 2012, Xingjun and Ping, 2016). In a research, Kashaninejad et al. (2009) found that the ANN modeling based on MLP network was better than the generalized mathematical Page model for prediction of moisture ratio. Subsequently, Kashiri et al. (2012) confirmed this result for modeling of sorghum hydration.

Generally, published literature is lacking a comprehensive study of hydration characteristics of wheat kernel and intelligent modeling of its hydration behavior using ANFIS model. Therefore, this study was carried out to develop an ANFIS intelligent model for directly prediction of moisture content, moisture ratio and hydration rate of wheat kernel during soaking based on simultaneous change of hydration temperature and time. The results were also compared with those obtained from ANN intelligent model. Moreover, the significance of the effect of hydration temperature and time on hydration behavior of wheat kernel was determined by means of statistical analysis methods.

Section snippets

Sample preparation

Shiroudi variety of wheat (Triticum spp.) was provided by Seed and Plant Breeding Unit, Agricultural Research Center of Fars province, Iran. The variety was selected, because it is one of the most commonly cultivated varieties in south region of country. Prior to hydration experiments, the samples were manually cleaned using napkin in order to eliminate foreign materials such as gravel, dust and broken kernels.

The initial moisture content of cleaned samples was determined according to ASABE

Physical properties

The length, width, thickness, mass, GMD, surface area, sphericity and vitreousness of experimental wheat variety were 7.042 ± 1.016 (mm), 3.231 ± 0.325 (mm), 2.462 ± 0.104 (mm), 0.032 ± 0.016 (g), 3.921 ± 0.374 (mm), 46.251 ± 0.561 (mm2), 60.136 ± 5.982% and 86.123 ± 6.108%, respectively.

Data analysis

Table 1 presents ANOVA results for water absorption of wheat kernel as affected by hydration time and temperature (main treatments). It can be seen in the table that the effect of main treatments and their interaction on water

Conclusions

Study and intelligent simulation of hydration characteristics (moisture content, moisture ratio and hydration rate) of wheat kernel during soaking resulted in achievement of the following conclusions:

  • 1.

    Statistical analysis results demonstrated that water absorption significantly increased with the increase of hydration time from 0 to 810 (min) or temperature from 30 to 70 (°C).

  • 2.

    Simultaneous effect of hydration time and temperature on water absorption was stronger than the effect of hydration time

Acknowledgment

Sincere thanks are stated to the reviewers for their invaluable comments on the manuscript.

References (84)

  • P. Garcia-Pascual et al.

    Morchella esculenta (morel) rehydration process modelling

    J. Food Eng.

    (2006)
  • M.A. Ghoush et al.

    Formulation and fuzzy modeling of emulsion stability and viscosity of a gum-protein emulsifier in a model mayonnaise system

    J. Food Eng.

    (2008)
  • W. Grzesiak et al.

    Methods of predicting milk yield in dairy cows-predictive capabilities of Wood’s lactation curve and artificial neural networks (ANNs)

    Comput. Electron. Agric.

    (2006)
  • A. Ibarz et al.

    Kinetic models for water adsorption and cooking time in chickpea soaked and treated by high pressure

    J. Food Eng.

    (2004)
  • E.A. Ibrahim

    Seed priming to alleviate salinity stress in germinating seeds

    J. Plant Physiol.

    (2016)
  • F. Jian et al.

    Temperature fluctuations and moisture migration in wheat stored for 15 months in a metal silo in Canada

    J. Stored Prod. Res.

    (2009)
  • V.A. Jideani et al.

    Modeling of water absorption of Botswana bambara varieties using Peleg’s equation

    J. Food Eng.

    (2009)
  • K.G. Kaptso et al.

    Physical properties and rehydration kinetics of two varieties of cowpea (Vigna unguiculata) and bambara groundnuts (Voandzeia subterranea) seeds

    J. Food Eng.

    (2008)
  • M. Kashaninejad et al.

    Modeling of wheat soaking using two artificial neural networks (MLP and RBF)

    J. Food Eng.

    (2009)
  • M. Kashaninejad et al.

    Study of hydration kinetics and density changes of rice (Tarom Mahali) during hydrothermal processing

    J. Food Eng.

    (2007)
  • J. Khazaei et al.

    Effect of temperature on hydration kinetics of sesame seeds (Sesamum indicum L.)

    J. Food Eng.

    (2009)
  • A.P. Kominakis et al.

    A preliminary study of the application of artificial neural networks to prediction of milk yield in dairy sheep

    Comput. Electron. Agric.

    (2002)
  • S. Lertworasirikul et al.

    Artificial neural network modeling of mass transfer during osmotic dehydration of kaffir lime peel

    J. Food Eng.

    (2010)
  • M. Maskan

    Effect of maturation and processing on water uptake characteristics of wheat

    J. Food Eng.

    (2001)
  • M. Maskan

    Effect of processing on hydration kinetics of three wheat products of the same variety

    J. Food Eng.

    (2002)
  • A.C. Miano et al.

    Correlation between morphology, hydration kinetics and mathematical models on Andean lupin (Lupinus mutabilis Sweet) grains

    LWT-Food Sci. Technol.

    (2015)
  • K. Mollazade et al.

    Analysis of texture-based features for predicting mechanical properties of horticultural products by laser light backscattering imaging

    Comput. Electron. Agric.

    (2013)
  • Y. Muramatsu et al.

    Volume changes of wheat and barley soaking in water

    J. Food Eng.

    (2006)
  • P. Oli et al.

    Parboiled rice: understanding from a materials science approach

    J. Food Eng.

    (2014)
  • A.L. Oliveira et al.

    Modelling the effect of temperature on the hydration kinetic of adzuki beans (Vigna angularis)

    J. Food Eng.

    (2013)
  • D. Petkovic et al.

    Determination of the most influential weather parameters on reference evapotranspiration by adaptive neuro-fuzzy methodology

    Comput. Electron. Agric.

    (2015)
  • A.C. Resio et al.

    Hydration kinetics of amaranth grain

    J. Food Eng.

    (2006)
  • M.H. Saeidirad et al.

    Predictions of viscoelastic behavior of pomegranate using artificial neural network and Maxwell model

    Comput. Electron. Agric.

    (2013)
  • K. Schulze et al.

    Development and assessment of different modeling approaches for size-mass estimation of mango fruits (Mangifera indica L., cv. ‘Nam Dokmai’)

    Comput. Electron. Agric.

    (2015)
  • P. Sefeedpari et al.

    Modeling output energy based on fossil fuels and electricity energy consumption on dairy farms of Iran: application of adaptive neural-fuzzy inference system technique

    Comput. Electron. Agric.

    (2014)
  • S.M. Shafaei et al.

    Analysis of water absorption of bean and chickpea during soaking using Peleg model

    J. Saudi Soc. Agric. Sci.

    (2016)
  • J.S. Torrecilla et al.

    Linear and non-linear modeling to identify vinegars in blends through spectroscopic data

    LWT-Food Sci. Technol.

    (2016)
  • M. Turhan et al.

    Application of Peleg model to study water absorption in chickpea during soaking

    J. Food Eng.

    (2002)
  • A. Vega-Galvez et al.

    Mathematical modelling of mass transfer during rehydration process of Aloe vera (Aloe barbadensis Miller)

    Food Bioprod. Process.

    (2009)
  • P. Wang et al.

    Physicochemical alterations of wheat gluten proteins upon dough formation and frozen storage – a review from gluten, glutenin and gliadin perspectives

    Trends Food Sci. Technol.

    (2015)
  • L. Yang et al.

    A prediction model for population occurrence of paddy stem borer (Scirpophaga incertulas), based on back propagation artificial neural network and principal components analysis

    Comput. Electron. Agric.

    (2009)
  • N. Abu-Ghannan et al.

    Hydration kinetics of kidney beans (Phaseulus vulgaris L.)

    J. Food Sci.

    (1997)
  • Cited by (39)

    View all citing articles on Scopus
    View full text