Abstract: With the continuous growth of the proportion of steel structure buildings in China, the proportion of steel frame structure buildings is also increasing year by year. Traditional design methods are no longer suitable for the requirements of high-quality development in China today. Therefore, this article uses the emerging Dung Beetle Optimization Algorithm to optimize steel frame structures. In response to the possibility of slow or even difficult convergence of the Dung Beetle Optimization Algorithm in complex problems such as building structure optimization, this paper proposes a method of using structural performance indicators (k) to control optimization to improve the Dung Beetle Optimization Algorithm. This greatly accelerates the optimization speed of the Dung Beetle Optimization Algorithm in steel frame structure optimization. Then, the best point set method is added to improve the population initialization stage of the Dung Beetle Algorithm, helping the algorithm more efficiently jump out of local optima and improve global optimization ability. This article uses Abaqus finite element analysis software to optimize the structure of a three-dimensional steel frame structure. The optimization results show that compared to the initial design, the proposed method can effectively reduce weight by 1 351.36kg, with an optimization ratio of 48.37%. This example demonstrates the efficiency and feasibility of the proposed method in optimizing steel frame structures.