Previous studies on demand-responsive buses rarely considered the impact of time-varying road networks and carbon emissions on vehicle scheduling, indicating the need for improvement in the limitations of existing studies. In response to the current scenario of mixed operation involving traditional fuel buses and electric buses under the backdrop of "dual-carbon, " this study outlines constraints, costs, and methods for measuring carbon emissions based on the characteristics of these two types of buses. It establishes a scheduling optimization model that incorporates delay time, carbon emissions, and operational costs as optimization objectives. To solve this model, it proposes the use of an adaptive genetic-firefly algorithm. The experimental results show that: (a) The proposed algorithm addresses the issue of local optimality common in traditional genetic algorithms. In experiments based on a simulated road network, it achieves a 9.1% reduction in the objective function, along with decreases of 0.3 vehicles, 4.9 nodes, and 104.57 km in average vehicle usage, average route nodes, and average travel distance, respectively, enhancing the precision of the solution. (b) Considering the impact of carbon emissions, the model can achieve a maximum reduction of 9% in carbon emissions and a 2.9% reduction in operating costs. (c) The vehicle scheduling scheme under dynamic impedance is both realistic and achieves simultaneous reductions of 7.5% in carbon emissions and 5% in operating costs. (d) The introduction of electric buses yields a significant reduction in carbon emissions, but the associated cost increase is noteworthy.

[2023-12-29] Accepted Paper /doc/2023.10.0506/v1