Addressing the issues of low initial exploration efficiency, slow convergence speed, poor solution quality, and susceptibility to local optima in traditional ant colony algorithms for multi-target point path planning, an adaptive ant colony algorithm tailored for multi-target point path planning is proposed. This algorithm is validated in the context of hedge trimming path planning. Firstly, by introducing the pheromone factor α, which increases based on the sine function, and the heuristic factor β, which decreases based on the cosine function, combined with the iteration process proportional coefficient to achieve dynamic parameter adjustment; Secondly, to address the problem of local optima, a triggered local optimization mechanism is further designed, where path nodes are randomly selected for sub-path re-planning after every M iterations, and on this basis, a dynamic node spacing strategy is introduced to balance global exploration and local exploitation capabilities. The results show that in 20×20 grid environments, 30×30 grid environments, and three-dimensional reconstruction of Hedge Path Planning, this algorithm has significant advantages in both path length and the number of iterations, indicating that it can effectively shorten path length and accelerate convergence in complex multi-target path planning tasks.