机场飞行区无人驾驶清水车优化调度方法

针对机场航班延误和拥堵现象日益严重以及地面特种车辆服务航班效率低且存在较高安全隐患的问题，研究了面向机场飞行区无人驾驶清水车的优化调度方法。通过将无人驾驶清水车服务航班硬时间窗与梯形模糊隶属度函数相结合构建航班服务水平函数，结合传统C-W节约算法，考虑无人驾驶清水车服务机场航班的时间规则，实现了无人驾驶清水车总行驶路程最短以及航班服务水平最高的目标。考虑服务航班数量总和，衡量每辆无人驾驶清水车的服务航班阈值，并提出了服务航班任务量的差异评价值。新算法在C-W节约算法路径优化结果的基础上对未达到服务航班容量极限的子路径进一步优化，实现了所需服务航班的无人驾驶清水车数量最少、服务航班数量差异化最小的目标。以国内某机场航班信息为例，结果表明:与单车单服务模式相比，服务总路程节省59.36%，车辆使用减少84车次，航班服务水平为93.78%，航班任务量的差异评价值由93.32%降低至43.96%；与基准算法相比，新算法在实现任务量均衡的同时并不会增加总行驶路程，且将服务航班任务量的差异评价值由2.72降低至0.44，显著提高了车辆服务航班任务量的均衡性。 

An Optimization Method for Scheduling Autonomous Potable Water Service Vehicles at Airfields

Due to increasingly serious flight delay and congestion and the issues of a low level of service and potential role of safety hazards of special vehicles at airports, an optimization method for scheduling autonomous potable water service (APWS) vehicles at airfields is studied. The level of service function for flights is developed by combining the hard time window of flights with a trapezoidal fuzzy membership function. Combined with the traditional C-W saving algorithm, the level of service function considers the time required for APWS vehicles serving flight, and with an objective to achieve the shortest total driving distance and the highest level of service to flights. Then, the total number of the flights to be served is used to measure the amount of work of each APWS vehicles, and an evaluation score for the amount of service work is proposed. Based on optimization results of C-W saving algorithm, the proposed algorithm further optimizes the sub-paths that do not reach the capacity limit of service flights, so as to achieve the minimum number of APWS vehicles and minimizing the difference in the number of flights served. A case study is carried out at a domestic airport, the results show that compared with the scenario with a single vehicle and uncoordinated service to flights, the total traveling distance of APWS vehicles is saved by 59.36%, 84 vehicle trips are saved, the level of service to flights reaches to 93.78%, and the difference of evaluation scores for the amount of service work is reduced from 93.32% to 43.96%. In contrast to the baseline algorithm, the workload of APWS vehicles can be balanced without increasing the total traveling distance, and the difference of evaluation scores for the amount of service work is reduced from 2.72 to 0.44, which significantly improves the workload balance of APWS vehicles.