A Dual-Mode Genetic-Ant Colony Algorithm for Dynamic Path Planning in Smart Scenic Spots

Abstract

This article focuses on the path planning problem in intelligent tourism scenic areas and innovatively proposes a dual-mode Genetic-Ant Colony Optimization algorithm (GA-ACO). This algorithm cleverly combines the global exploration advantage of genetic algorithms (GA) with the local development expertise of ant colony optimization (ACO), constructing a multi-dimensional dynamic fitness function in real-time that incorporates both visitor preferences and dynamic environmental constraints (such as attraction congestion index and opening status), thereby achieving refined evaluation of path quality. The algorithm facilitates deep collaboration between GA and ACO through the design of pheromone-guided genetic operators (which prioritize high-quality path segments with high pheromone concentrations during selection, crossover, and mutation operations) and a bidirectional feedback mechanism (where elite solutions generated by GA are converted into the initial pheromone matrix for ACO, providing guidance for multi-modal search; the pheromone accumulated by ACO provides local heuristic information for GA), forming a closed-loop enhancement of global exploration and local development. Experiments show that, compared to traditional algorithms (such as pure GA and ACO), GA-ACO significantly outperforms key indicators like path length, number of iterations, and number of turns: the optimal path length is reduced by 6% compared to ACO, the average number of iterations is reduced by 50%, and the number of turns is minimal (only 3.8 turns); compared to GA, path smoothness and energy efficiency are significantly improved. Additionally, it can accurately generate optimal paths in three-dimensional space simulations, with reliability verification showing a prediction accuracy of 99.26%, and it has significant advantages in solving the total number of feasible paths and the time taken to find the optimal path, allowing for rapid responses to dynamic changes in scenic areas, thereby providing efficient and reliable solutions for path planning in intelligent scenic areas.