Attiéké and fermented cassava dough are two cassava derivatives linked by the same transformation process. Fermented dough is obtained after spontaneous fermentation of the pulp over 2 to 3 days, with the addition of inoculum on cassava. Fermentation of dough for attiéké occupy an important role in the availability of fast sugars, mineral salts and cyanide, and in granulation. The dough becomes more acid under the action of Yeasts, Molds, Bacillus and mainly Lactobacillus. Dehydrated by pressing, it is generally characterized by a pH around 4.5, a humidity of around 52% for a reducing sugar content of around 1%. The aspect of the fermented dough that precedes attiéké depends on the factors and conditions involved in fermentation, the force of pressing and the hygiene of operations. However, the traditional processes and properties of attiéké have been assessed without prior studying the technological and biochemical properties of fermented cassava dough.

The present study consists to using artificial neural networks to create mathematical models allowing to predict the growth of tomato plants and to compare them to the growth in real time in order to control the productivity of the tomato. Tomato growth was modeled by an empirical model using artificial neural networks as a tool through a program developed in the Matlab R2010b software. Mathematical models were developed to predict the growth of the tomato plant for the number of leaves, leaf length and width, height and circumference of the plant. The experiments were carried out in the regions of High Sassandra (Daloa, Côte d’Ivoire). The coefficients of determination between the experimental measurements and the measurements predicted by artificial neural networks are respectively 0.9722; 0.9925; 0.997; 0.9945 and 0.9926 for plant height; the number of sheets; the circumference of the plant; leaf length and leaf width. These results are satisfactory insofar as all the coefficients of determination (R2) are greater than 0.97. Likewise, the curves representing the predicted values and the experimental values have practically the same appearances or even confused. These results show a good interpolation between the experimental values and those predicted by the mathematical models.

This study aimed to describe the behavior of ginger, and to predict its water content, during artificial drying under four temperatures (60 oC, 80 oC, 100 oC, 120 oC). Experiments were carried out on ginger using a DRY-Line type oven. The obtained data was fitted using 4 semi-empirical thin layer drying models. Among the semi-empirical models considered, the diffusion approach model was chosen as the most appropriate model to describe the behavior of ginger. For the different temperatures, he presented respectively the coefficients (r) of 0.9970; 0.9974; 0.9949 and 0.9942; the coefficients Chi-square (χ²) of 4.0306 X 10-6, 3.7015 X 10-7; 1.6387 X 10-7 and 1.3637 X 10-6 and Root Mean Square Errors (RMSE) of 3.5851 X 10-4; 1.1415 X 10-4; 9.2226 X 10-5 and 2.6604 X 10-4 for the four temperatures. The diffusion coefficient varies from 9.585 X 10-9 to 3.466 X 10-8 m2/s and strongly depends on the drying temperature. The activation energy is estimated at 24.188 kJ/mol.

The tomato is an annual herbaceous plant, of the Solanaceae family. It is cultivated for its fruits which are consumed either fresh or cooked, or processed industrially. Its growth is a complex phenomenon which involves several parameters. A study of the growth parameters carried out in the region of Daloa (Côte d’Ivoire) showed a complexity of the growth of the tomato at the level of the number of leaves, the length of the leaves, the width of the leaves, the height of the trunk and the circumference of the trunk of the tomato plant. For this purpose, mathematical models were developed to predict the growth of the tomato plant from artificial neural networks for the number of leaves, the length of the leaves, the width of the leaves, the height of the plant and the circumference of the trunk of the tomato plant. The coefficients of determination between the experimental measurements and the measurements predicted by artificial neural networks are respectively 0.9722; 0.9925; 0.997; 0.9945 and 0.9926 for plant height, number of leaves, plant circumference, leaf length and width. These results are satisfactory insofar as all the coefficients of determination (R2) are greater than 0.97. These coefficients close to 1 show a good interpolation between the experimental values ​​and those predicted by the model. They indicate that the values ​​predicted by artificial neural networks are almost more than 97% close to the experimental values. Because of this, artificial neural networks are reliable enough to predict tomato growth in leaf count, leaf length, leaf width, plant height, and trunk circumference of the tomato plant.

The aim of this study is to compare a greenhouse solar drying simulated by a developed numerical model, with the actual experimental drying of cocoa beans. Thus, using multiphysical models (based on conservation laws of energy, mass and momentum within the dryer and beans), it was possible to simulate the greenhouse drying behavior of the beans. The resolution of the developed equation system was done using the finite element method of COMSOL Multiphysics 4.0 software. For the validation of the numerical model, an experimental study has been developed at the designed dryer. In this study, a drying operation of cocoa beans was conducted. The characteristics of the drying air (temperature, relative humidity and speed) as well as the mass of the beans were regularly determined over time. The results indicate that, in general, there is good agreement between the experimental results and the simulated results (R close to unity). This allows a validation of the mathematical model of the drying of cocoa beans within the greenhouse type solar dryer used.