An Improved Multi-objective Evolutionary Algorithm Based on Gaussian Distribution Estimation

When an emergency occurs, how to specify a reasonable resource scheduling scheme significantly affects disaster relief efficiency. However, most actual existing schemes lack considering satisfaction of potential disaster sites, and lack a scheduling model with 3 or more optimization goals, which makes it difficult to apply to complex scenarios. In this paper, we propose a four-objective resource scheduling optimization model that additionally considers potential disaster sites satisfaction. And we have designed an improved NSGA-III-GD algorithm to optimize this model. First, we introduce NSGA-III, an algorithm that has a great advantage in multi-objective optimization problems. And more importantly, we use Gaussian estimation distribution instead of traditional cross mutation operators to extract the overall characteristics of the population, which improves the search accuracy of the optimal solution and greatly improves the convergence speed. The experimental results clearly show that the algorithm proposed in this paper has achieved very good performance.