Abstract:

To reduce casualties and economic losses caused by disasters, a multi-objective ambulance dispatch optimization model that minimizes the maximum rescue time, the standard deviation of ambulance service times, and the total psychological deprivation cost of casualties was developed, while accounting for psychological deprivation factors. Given the non-deterministic polynomial (NP)-hard nature of model, the INSGA-III was proposed, incorporating a multi-chromosome hierarchical encoding strategy and dynamic crossover and mutation methods. Using the 2019 flood in Xingguo County, Ganzhou City, Jiangxi Province as a case study, the proposed INSGA-III was compared with NSGA-III and NSGA-II. Sensitivity analyses on ambulance quantity and relative deprivation cost coefficients demonstrated the effectiveness of model and algorithm. The results show that the maximum rescue time for casualties is minimized to 9.234 hours, the standard deviation of the latest ambulance service times is reduced to 13.156 minutes, and the total psychological deprivation cost of casualties is minimized to 1729.001. With the relative psychological deprivation cost coefficient set at 0.3 and 500 ambulances deployed, the timeliness and fairness of the rescue operations are significantly improved.

Key words: flood disaster, psychology of injured, ambulance, multi-objective optimization, emergency rescue, improved third-generation non-dominated sorting genetic algorithm (INSGA-Ⅲ)