Multi-Objective Programming-Based Tour Route Planning

Abstract

This paper develops personalized 144-hour itineraries for foreign tourists in China, accounting for differences in preferences, attraction density, travel time, and the combined cost of admission and transportation. For Problem 1, multi-source appendix data were merged into a single Excel sheet and attractions with missing ratings were removed. Ratings were then sorted, showing 5 as the maximum score; 2,563 attractions achieved this level, and cities were grouped by their counts of 5-point (BS) attractions. Among 334 cities with BS attractions, 16 cities have at least 15 such sites, with the top ten including Sansha, Wujiaqu, Yuxi, Yiyang, Tianmen, Yantai, Weifang, Greater Khingan Range, Alar, and Xingtai.

For Question 2, an evaluation system covering city scale, environmental protection, cultural heritage, and transport convenience was built. After standardization, indicator weights were derived via the entropy method and combined with TOPSIS to rank cities; SPSS produced the “Top 50 Most Desirable Cities for Foreign Tourists.”

For Questions 3–5, attraction clustering along the southeast coast was used to reduce travel time, and multi-objective integer programming models were solved in Matlab. The resulting itineraries cover 14 cities (124.8 hours, 1,812 yuan), 13 cities under stronger cost constraints (107.38 hours, 868 yuan), and a mountain-themed 10-city route (111.95 hours, 1,443 yuan).