Based on the circulation transportation network, this study establishes the model of tractor optimal op- eration and designs an improved genetic algorithm according to the heuristic rules to solve. Numerical examples illustrate the effectiveness and feasibility of the proposed model and algorithm. From the comparison, we know that the improved genetic algorithm based on heuristic rules is better than the solving strategy of simulated an- nealing algorithm with the efficiency of the tractor to complete all tasks increased by 5.54%. The simulation re- sults show that the proposed method possesses high convergence speed and higher quality solutions, which may provide significant guidance for the practical application of scheduling of semi-trailer swap transportation.