Abstract: A scheduling and transportation path planning method based on improved NSGA-II is proposed to address the problems of low scheduling and path planning efficiency of existing intelligent trackless auxiliary transportation systems in coal mines, as well as insufficient consideration of practical constraints. A multi-objective optimization function for the intelligent trackless auxiliary transportation system in coal mines is established with minimum transportation cost, minimum total waiting time of electric locomotive robots, and minimum expected deviation of coal mine transportation as objective functions, and constraints such as minimum unloading volume, total capacity of loading points, loading frequency, and maximum unloading capacity of unloading points. A multi-objective optimization algorithm for improving NSGA-II is proposed, which utilizes strategies such as Levy flight, random walk, and adaptive weights to enhance the global and local search capabilities of the algorithm and accelerate its convergence speed. Simulation experiments show that compared with the no optimization scheme, the improved NSGA-II optimized balance scheme reduces transportation costs by about 19%, shortens queuing waiting time by about 56%, and reduces the minimum expected deviation of coal mine transportation by about 40.5%. The experimental results have verified the effectiveness and practicality of the proposed improved NSGA-II algorithm optimization results, which can provide multiple optimization options for underground production management in coal mines and have broad application prospects.