<p>As the resources of open-pit mines diminish, deep mining increases transportation costs, making transportation a critical aspect of mine production. Designing an efficient transportation system to reduce these costs is essential, with route planning being its core. However, traditional algorithms struggle with adaptability in the complex environment of mines. This study introduces an improved Sparrow Search Algorithm (SSA) and a two-layer model to optimize the route planning for mine transport vehicles. The study conducts a series of simulation experiments. Using a grid method to convert environmental information into discrete grid units, the experiment is based on the complex terrain of a real mine. It sets slope penalty costs and considers multiple feasible routes. The performance of the improved SSA is tested using the double-peak functions F12 and F13, compared with traditional SSA, Grey Wolf Optimizer (GWO), and Whale Optimization Algorithm (WOA). Additionally, the study simulates sudden dynamic events on the mine’s transportation roads, comparing them with traditional static models and other SSA variants in large-scale open-pit mines. The results show that the improved SSA performes better in the double-peak function tests, and the two-layer path planning model excels in dynamic scenarios. The research method outperforms the comparison methods in terms of path length, planning time, energy consumption, and feasibility frequency. In practical applications, it can significantly reduce transportation costs, shorten time, reduce energy consumption, and improve task completion rates. However, the algorithm has weak adaptability to complex dynamic road conditions. In the future, real-time perception technology can be combined to improve real-time adaptability and robustness.</p>