Abstract: The visitor's recreational experience is a multi-stage experience process, including anticipation, on-site travel, on-site participation, departure from the site, and recall. At present, the entire tourism industry and related infrastructure are undergoing digitization, and tourism digitization is called e-tourism. In the future, e-tourism will extend to the entire travel cycle, and nature tourism in protected areas is no exception. This article mainly adopt sinter disciplinary exploratory research methods such as literature review, field investigation, expert scoring, and mathematical modeling to innovate the application model of the visitor recreation management tool in the traditional protected area—Tourism Optimization Management Model (TOMM theory). Taking the Three-River-Source National Park as a case, we use artificial intelligence algorithms to solve the core problem of visitors with different recreational needs—"customized optimal route planning with dynamic experience". With this as the core, it is possible to build a practical scenario for the application of TOMM theory in the context of e-tourism, that is, to realize the application of TOMM theory in the national park application scenario. TOMM theory is based on the application mode of "customized optimal route planning with dynamic experience" APP in the context of smart national parks, which will solve the two major practical difficulties in traditional TOMM theory, that is, it is difficult to control the way and degree of public active participation, and it is difficult to obtain and monitor real-time data streams. At the same time, the research will provide a research foundation for the establishment of the smart protected area system with Chinese characteristics.