Product pipelines, serving downstream markets, need to adopt a scheduling mode based on users’ demand. So far,
some research has focused on the optimization of product pipeline scheduling, but little has considered the demanding time-windows,
which often lead to the dissatisfaction of users. Aiming at single-source and multi-distribution product pipelines, this
paper sets the minimum deviation between the actual delivery time-windows and users’ required time-windows as an objective
function and proposed a MILP model solved through the ant colony algorithm and the branch and bound algorithm, which take
the constraints of batches, injection, delivery and time nodes into consideration. Then the paper chose one of product pipelines
in China as an example, then modeled and solved the detailed scheduling plan. The results demonstrate the strong efficiency and
practicability of the model in solving the detailed scheduling plan with only a small deviation.
Key words：
product pipelines; optimization of scheduling plan; demanding time-window; MILP model; branch and bound; ant colony optimization algorithm
