Written Paper

Optimization of petfood using mathematical modeling  [2015]

Alexeev, G.V. Aksenova, O.I., St. Petersburg State National Reseach Univ. of Information Technology, Optical Design and Engineering (Russian Federation) Derkanosova, A.A., Voronezh State Univ. of Engineering Technologies (Russian Federation)

Visualise le texte intégral

Proposed is a new approach to increasing in sphere provender production for domestic animal based on the use of secondary feedstock. Optimum formulations were developed by using mathematical modeling based on device a fuzzy logic apparatus. The advantage of the given method is a possibility of undertaking an analysis, in conditions of the subjective estimation of empirical data and the absence of the evident numeric form of the result. For realization of the proposed methods the Fuzzy Logic Toolbox module, belonging within the MatLab package is used, but for presentation of results of modeling in graphic form, the Surfase Viewer module is used. As input variable ranges there were used pH (5-8); moisture content (7-11%); concentration (50-90%); particle size (0.3-0.8 mm); food value (100-500 kcal/100 g); overall biological bacterial content (0-1000 cells in 1 g of food). To supply the input rules to be included into the system after building membership functions, values of input variable meeting the best quality attributes were set. Evaluating the palatability by animals has found that its best organoleptic characteristics are achieved under the following values of the main parameters: pH is 6.5; moisture content is 9%; concentration of protein constituent is 85%; particle size is 0.55 mm; food value is 267 kcal/100 g. Several formulations of extruded dry feeds were developed wherein optimal concentrations of protein component and moisture content were dete
rmined. Modeling by classic statistic methods and fuzzy logic methods gave close results. The fuzzy logic methods are invited to use in modeling feeds for young animals.


Proceedings of the Voronezh State University of Engineering Technologies

ISSN : 2226-910X