The work aims to propose a hybrid algorithm of tower loading heuristic algorithm, two-dimensional loading point heuristic algorithm, and ant colony simulated annealing algorithm to address the container loading problem of large-scale and complex multi specification goods in the logistics industry. Firstly, the three-dimensional container was loaded into towers through the tower loading heuristic algorithm to reduce the three-dimensional packing problem to a two-dimensional packing problem, effectively reducing the loading scale of large-scale containers. Secondly, the ant colony algorithm incorporates a pheromone selection and update strategy and an adaptive pheromone evaporation coefficient to improve the overall convergence speed of the algorithm. At the same time, it combines with simulated annealing algorithm to perform local search on the set of excellent paths in each generation, avoiding the algorithm from falling into local optima due to too fast convergence. Finally, the ant colony simulated annealing algorithm was combined with a two-dimensional loading point heuristic algorithm to optimize the loading sequence and placement posture of each tower to find the optimal loading plan. The experiment showed that in 250 sets of examples, the average space utilization rate of the container in this algorithm was 90.92%, which was better than that of the three comparative algorithms. In conclusion, the hybrid ant colony simulated annealing algorithm designed in this article is very suitable for solving large-scale container packing problems.