The purpose of this study is to explore the optimal regional epidemic isolation pattern by optimizing the locations of emergency resource centers. To aim at minimizing the distance, a genetic algorithm is used to obtain the global distances under isolation patterns by administrative division, k-means clustering and grid partition, respectively. The results show that the global distance of the isolation by administrative division is the longest while the global distance of the isolation by k-means clustering is the shortest. Under the isolation by the k-means clustering or the grid partition, the global distance can be shortened if there are more partitions. During the epidemic isolation, emergency resource centers in the central or on the edge of an area should be constructed differently. Emergency demand points on the edge of an administrative district should be divided by the nearest distance. There could be delicacy partitions within the isolated areas for better management.

epidemic isolation  /  emergency resource  /  emergency logistics  /  genetic algorithm