The design of the grain supply chain network is one of the important and challenging problems in the field of agri-food supply chain. Almost all researches mainly concerned with the problem in deterministic, stochastic or fuzzy environments. Different from them, we study the grain supply chain design problem by using uncertainty theory. This paper employs an uncertain programming tool to design the grain supply chain network and a α-cost minimization model is proposed. To solve the model, we design a hybrid intelligent algorithm which integrates 99-method and different evolution algorithm. Finally, a numerical example is presented to illustrate the effectiveness of the proposed model and algorithm.