Thematic cultural heritage tourism trail planning integrating multi-source data and machine learning in Wuhan China

Nowadays, Cultural heritage tourism faces challenges in route planning, including weak data-mining capacity, limited multi-indicator evaluation, and inefficiencies in traditional pathfinding. This study proposes an innovative thematic and sustainable framework that integrates advanced digital technologies at both meso- and micro-spatial scales to optimize heritage route planning. The study introduces and applies the Non-dominated Sorting Genetic Algorithm III (NSGA-III)—specifically designed for high-dimensional multi-objective optimization—which outperforms existing methods in key aspects and effectively solves complex route optimization problems under multiple constraints. Experimental results confirm that Nsga3ip demonstrating 97% rational route probability and 0.89 optimization efficiency—surpassing MOPSO (83%, 0.62) and random algorithms (12%, 0.19) under identical constraints. The findings demonstrate its strengths in planning quality, enhancement of heritage value, and practicability. This underscores the method’s innovation and applicability, further promoting the integration of data-driven approaches in heritage conservation and interdisciplinary urban research.