The traditional path planning algorithms that combine ant colony optimization (ACO) and genetic algorithm (GA) commonly suffer from unsmooth paths, slow convergence speed and high energy consumption. To address these issues, a path planning method based on dynamic fusion of ACO and GA (DACO-GA) is proposed to improve the efficiency and accuracy of path planning. In the initial stage, ACO was used to generate the initial population, and GA was introduced for optimization and adjustment. In later stages, the leading role of the two algorithms was dynamically switched, enabling coordinated complementarity between global and local search. The algorithm integrated adaptive pheromone distribution, dynamic evaporation factors and adaptive crossover/mutation probability adjustment mechanisms, which effectively enhanced search capability and mitigate the tendency to fall into local optima. Finally, optimization experiments were conducted on the key control parameters of the DACO-GA to validate the effectiveness of each improvement mechanism. The DACO-GA was compared with traditional algorithms across multiple typical scenarios to further evaluate its adaptability in complex environments. The results showed that the proposed algorithm could generate smoother and shorter paths, demonstrating strong global optimization ability and faster convergence speed. The DACO-GA not only provides an effective solution for complex path planning problems, but also offers technical references for the optimization in areas such as multi-agent cooperation and robot navigation.