Abstract: In order to further improve the solution quality and optimization effectiveness of the short-term crude oil scheduling problem, a two-stage optimization strategy for addressing such problems is proposed. First, Through the analysis of the assignment process from charging tanks to distillers, a crossover operator that can preserve segmentally parent genes and a mutation operator that adaptively changes mutation probabilities are given. Additionally, the NSGA-III-ACMO algorithm is introduced to solve the short-term crude oil scheduling problem, which ensures good convergence and population diversity while optimizing five objectives: crude oil mixing cost in pipeline and in charging tanks, tank-switching cost in distillers, tank usage cost, and energy consumption cost. To address the issue of incomplete optimization of energy consumption cost, a new mixed integer linear programming model is proposed for further optimization. The advantage of this model is that, for a given detailed schedule, it can minimize the energy consumption without affecting other objectives. A case study demonstrates that comparing the schedule obtained by the NSGA-III-ACMO algorithm with the results of existing literature, the optimization of individual objectives is improved by 9% to 45%. On this basis, the proposed model can further reduce energy consumption cost by 6.8%. Overall, the NSGA-III-ACMO shows the obvious superiority in both solution quality and optimization effectiveness.