To enhance the flexibility of electricity-heat-hydrogen systems and reduce carbon emissions, an optimal scheduling strategy for hydrogen-integrated multi-modal utilization systems was developed considering flexibility constraints and an improved stepped carbon trading mechanism. Models of hydrogen energy utilization equipment were established, and the regulation capabilities of hydrogen devices as well as flexibility resources on both the supply and demand sides were quantified. On this basis, a two-layer low-carbon economic dispatch model incorporating flexibility constraints was formulated. In the upper layer, an improved stepped carbon trading optimization model was developed, where a differential evolution algorithm was applied to achieve adaptive adjustment of trading parameters. In the lower layer, a dispatch simulation model considering flexibility constraints was constructed, with system operating cost set as the objective function and solved using the Cplex solver. Through iterative optimization between the two layers, the optimal scheduling scheme was obtained. Simulation results under three representative scenarios demonstrated that the proposed strategy effectively reduced carbon emissions while ensuring the balance of supply and demand flexibility within the system.