In response to the challenges posed by the shipping carbon tax resulting from the low-carbon environmental policy, the seasonal fluctuation of freight rates in the container liner shipping market, and the reduced flexibility in fleet adjustment, we develop a bi-level programming model to maximize the average revenue per container ship and the total revenue of the fleet.The upper-level model selects the optimal shipping networks for both the off-season and peak season, while also designing the liner fleet for these two networks. The lower-level model optimizes the slot allocation scheme and the empty container storage and transportation scheme to evaluate the revenue of the network calculated by the upper-level model. Three meta-heuristic algorithms are proposed. We take the China-West Europe liner shipping route as the research object, and conduct numerical experiments on different optimization objectives and liner types. Results demonstrate that maximizing the average revenue per container ship involves reducing the carbon tax cost, simplifying the trunk route structure, and increasing the number of feeder routes. These changes lead to reduced satisfaction rates for transportation demands in both China and Europe. If the company seeks to maximize total revenue, they may achieve the opposite result. Therefore, liner companies should reasonably set optimization goals, adjust the network structure and liner operation status in a timely manner, and scientifically allocate container ship and empty containers to achieve the operational goals of reducing carbon emissions and maximizing revenue to cope with the volatile market and the low-carbon regulations.