In the development of modern logistics, the timeliness and cost of multimodal transport are factors that cannot be ignored. Aiming at the transportation goal of highway-railway-waterway multimodal transport, this paper mainly studies the path optimization problem of green multimodal transport when the dual uncertainties of transportation time and transit time follow random distribution. The multi-objective optimization model of green multimodal transport path under the dual uncertainties of transportation time and transit time is established with transportation time, carbon emissions and transportation cost as objective functions and carbon emissions as constraints. Accordingly, fuzzy adaptive genetic algorithm and fast non-dominated sorting genetic algorithm (NSGA- Ⅱ) are used to design the multimodal transport path optimization strategy. Finally, the path data from Nanchang to Berlin is used to verify the effectiveness of the proposed method and the results are compared and analyzed. This paper finds that the multi-objective optimization results based on NSGA-Ⅱ algorithm are better, which can guide the multimodal operators to adjust the transportation plans and reduce the emission of carbon dioxide, providing reference for logistics enterprises to carry out multimodal transportation.