| Feed is material base of cultivation Industry. The feed formula has significant meaning in poultry raising industry. Feed manufacturers often adjust feed formulation because cost of raw material has been changes. The goal of adjusting feed Formulation is to lower cost as far as possible under the premise of guaranteeing feed's quality and nutrition. Because production of feed is large cost reduction can contribute to bring Considerable economic benefits even it is small-scale. Rapid calculation of least-cost formula and using in the production of feed is changing the role of formula design in today's feed industry.The mathematics method plays a very important role in feed formulation. How to use mathematical methods to design or optimize feed formulations to lower production costs and access to the lowest price have been a hot research problem under the conditions meeting the standards of economic animal husbandry or user requirements. In this thesis, four types of mathematical programming methods of linear programming, goal programming, stochastic programming, and fuzzy programming as well as two types of mathematical optimization algorithm simulated annealing algorithm and genetic algorithm is applied to the design of feed formulas.The formulas which are obtained to use various methods can be said to have their own advantages and disadvantages. selecting the specific formula is related for market conditions, feed costs and feed nutrition standards. In the essay meeting the feeding standards, the cost of feed formulations, nutritional balance, feeding effect of four areas will be as a judge the merits of feed formulas. Finally self-organizing neural network model is used to make decision about the quality of the various feed formulations. Designed an ideal formula, a middle formula and an out formula in the decision-making. Self-organizing feature map neural network (SOFM) possesses strong clustering feature. Finally according to the clustering results elect optimal formula. |
PDF Full Text Request