Optimization of Emergency Shelter Siting and Material Distribution Considering Multiple Psychological Perceptions of Disaster Victims Under Two-Dimensional Fairness Measurement

In the early stages of large-scale disasters, governments are tasked with temporary relocation of disaster victims and distribution of essential supplies. During this period, delays in emergency resource allocation and uneven distribution among shelters can lead to negative psychological impacts among disaster victims, thus affecting post-disaster rescue efficiency and psychological counseling efforts. To address this issue, this study constructs a cost function to measure the perceived suffering of disaster victims. From a two-dimensional (2-D) fairness perspective, it also formulates a comparison penalty cost function to quantify the envy effect arising from unequal distribution. In consideration of minimizing the overall social cost of rescue operations, the study develops an integrated optimization model for evacuation, shelter site selection, and material allocation that balances efficiency and fairness. Furthermore, an improved particle swarm optimization algorithm incorporating Taboo search (TSPSO) is designed for solution purposes. The model and algorithm are applied to a real case study during Typhoon “Doksuri” in 2023, which struck Quanzhou City, Fujian Province, along with four different randomly generated scenarios of varying scales. Comparative analysis against two classical algorithms is conducted. Case studies demonstrate that utilizing fairness coefficients and perceived suffering cost functions can achieve more equitable and effective humanitarian rescue operations. The results validate the effectiveness and feasibility of the model and algorithm, showing that TSPSO outperforms the other two classical algorithms.