Decision Rules for the Dynamic Animal Feeding Problem
用动态规划推导出基于动物体重的最优喂养和出售决策规则，能适应投入产出价格的短期随机变化，并引入一个涵盖无限序列和单批次模型的新随机模型。
Abstract Under a criterion of maximum expected present value, dynamic programming is used to derive sequential decision rules that give feed rations as a function of animal weight and indicate the critical weight at which to sell a batch of animals. These rules are adaptable to short‐term random changes in prices of inputs and outputs, and typically require solution of a single equation at a given weight. A new stochastic model is introduced that encompasses both the infinite sequence and single batch models found in the literature; the infinite chain of batches is randomly terminated after each batch is marketed.