Abstract: In the text of carbon peaking and carbon neutrality strategy and green shipping, the optimization of feeder transportation scheduling is studied for a fleet composed of container ships using different kind of propulsion power. A fleet composed of LNG powered ships and oil burning ships is studied as an example. A fuzzy time window is introduced to describe the degree of satisfaction of the consignor. Under the constraints of transportation capacity of the ships and the ports' time restriction, a mixed integer programming model is built, taking comprehensive transport costs, the satisfaction degree of consignor and carbon emissions into consideration. A hybrid genetic algorithm to solve the problem is designed based on the characteristics of the problem. The local search strategy is introduced to improve the solution quality, achieving, as test calculation shows, the deviation of 1.4% from optimal solution in average. The calculation with the mixed genetic algorithm converges faster than with conventional genetic algorithm and has stronger optimization search ability. Calculation shows that the optimization can reduce transport costs by 29 520 yuan and carbon emissions by 38 474 tons.