Optimization of Sightseeing Paths in Tourism Park Based on Improved Ant Colony Algorithm

Abstract

The intelligent navigation system of tourism parks typically functions to recommend diverse sightseeing paths for tourists of different age groups, comprehensively considering factors such as path distances and the geographical location of all the scenic spots. However, the classical ant colony algorithm often get stuck in local optima during path planning, alongside exhibiting low planning efficiency and a low convergence rate. This paper presents an improved ant colony algorithm combining the classical ant colony algorithm and the genetic algorithm, seeking to enhance the efficiency of the park navigation system in planning sightseeing paths. Initially, the algorithm generates initial pheromone distributions for identifying the optimal path, based on the genetic algorithm and through a strategy of crossover and mutation. Subsequently, the process entails distinct optimizations, utilizing the heuristic function, pheromone update mechanism, and state transition strategy within the ant colony algorithm, ultimately arriving at the optimal solution for the traveling salesman problem (TSP) and further refining paths. The experimental verification showed a path distance of 305.62 meters at the 87th convergence using the improved ACA+GA algorithm, with 250 repeated iterations at α=1.5, β=3, ρ=0.5, Q=300. In contrast, the classical ant colony algorithm resulted in a path distance of 519.74 meters by the 196th convergence. These results showcased the efficacy of the improved ant colony algorithm in overcoming the shortcomings of the classical ant colony algorithm, achieving a higher rate in reaching the optimal solution and an enhanced efficiency of path planning.

Keywords

• intelligent navigation system of tourism park
• traveling salesman problem(tsp)
• ant colony algorithm
• genetic algorithm