Abstract: Most crude oil scheduling schemes are compiled by personnel experience or mathematical programming methods in refinery industries. It is difficult to meet the requirements of the crude oil online blending or to reduce the scale of optimization by these methods in real practice. For the typical crude oil pipeline scheduling process, a receding horizon scheduling method was developed in which crude oil blending demands were considered. The mixed-integer programming was employed first to establish the crude oil scheduling model, and then the blending demands were incorporated into the constraints as well as the objective function. By dividing the whole scheduling period into several short ones, the receding horizon optimization was implemented to solve this problem. The results indicate that with the scheduling period of 168 h, the crude oil pipeline scheduling problem can be successfully resolved by the proposed method under the condition of dynamic crude oil blending formula. Obviously, this method can make the crude oil scheduling match the crude oil blending demand much better, and the more economical crude oil scheduling scheme can be found quickly, thereby greatly promoting the economic efficiency of the refineries.

Key words: crude oil online blending, crude oil pipeline scheduling, mixed-integer programming, receding horizon optimization