首页 > , Vol. , Issue () : -
As important resources for solving geo-problems, simulating geographic environments, and supporting decision-making analysis, geographic analysis models play an important role in the construction of virtual geographic environments (or digital twin geographic environments). Integrating geographical analysis models with different simulation capabilities in multiple fields can more comprehensively express the real geographical environment and support the construction of visual, usable, analyzable, and interactive virtual geographic environments. Aiming at integration methods of geographic analysis models, this paper systematically sorts out the current research progress, and summarizes the evolution characteristics of geographic analysis models, integration methods, and related supporting technologies of geographic information services. It is concluded that solving complex geographic problems through multi-model integration has become a widely recognized research approach, and integrated modeling that integrates nature-humanities and integrates big data and deep learning has also attracted more and more attention. However, the existing model integration methods still face the limitation of "case-by-case". It is difficult to form continuous and reusable integration results in dealing with the compatibility of heterogeneous models, which makes model integration easy to fall into repeated technical programming work. Drawing on the concept of containerization in the field of information services, this paper proposes the idea of "geographic analysis model container" and implements the methods. By designing the information interaction channel based on a flexible assembly structure and the functional computing channel based on the event response mechanism, a container that can continuously accumulate and improve the model integration and reuse capability is constructed, and the configurable integration application can be performed based on such model containers.