For the proportional flow-shop scheduling problem with multiple customer orders, due dates, earliness and tardiness penalties are taken into consideration. The cooperative games model is established with the customer's expenditure cost as the optimization objective, and the maximum cost savings obtained from the rescheduling as the worth of the coalition, which is formed by customers through cooperation. It is proved that the corresponding cooperative games which have no externality are balanced and have a nonempty core. Taking into account the different urgency degree of customers to advance processing and delay processing, a kind of cost allocation method based on the beta rule with earliness and tardiness penalties is proposed, and it is also proved that this allocation method gives a core allocation of the game. Finally, a hybrid differential evolutionary algorithm is designed to solve the optimal scheduling arrangement. The experimental results verify the effectiveness of the scheduling strategy and the cost allocation method based on the cooperative games model.