Goal programming with penalty functions and livestock ration formulation

Abstract This paper contends that the conventional linear programming paradigm commonly used for livestock ration formulation suffers from several inherent weaknesses. One such weakness is the mathematical rigidity that must be respected to meet the minimum nutritional requirements specified in the model, thus making the paradigm inflexible and unrealistic under many situations. Goal programming techniques, on the other hand, do not impose such rigid conditions and also allow consideration of several decision criteria. This paper represents a departure from the traditional LP approach by formulating the diet formulation problem as a GP model incorporating penalty functions that make the specification of minimum levels of nutrients more flexible and realistic.