关键词: 应急物流, 车辆路径优化问题, 多目标鲁棒优化, 改进SPEA2算法

Abstract: This paper aims at the vehicle scheduling requirements of economy, timeliness, reliability and robustness in emergency logistics. A multi-objective robust vehicle routing problem with time windows, uncertain de-mand, uncertain driving time and routing failure risk is considered. A new cost function, a satisfaction function, a risk function and a robustness function are proposed to be four optimization objectives of the model, and the uncertain model is transformed into a deterministic robust counterpart model based on the robust optimization theory. In this paper, the multi-objective model is solved based on the SPEA2 algorithm framework, but a variety of improvement strategies are proposed for the algorithm defects. The effectiveness of the improvement strategies is proved by comparison experiments.

Key words: emergency logistics, Vehicle Routing Problem（VRP）, multi-objective robust optimization, Improved Strength Pareto Evolutionary Algorithm 2（ISPEA2）