Under a traditional design scheme, the design quantity of the main girder of the bridge-erecting machine is redundant and there are many consumables, which reduces the production efficiency, and existing optimization methods have the problem of low convergence accuracy. Therefore, this paper proposes an improved beluga whale optimization algorithm based on the quadratic interpolation strategy and carries out the lightweight design of the main girder of the bridge-erecting machine. By introducing the quadratic interpolation strategy, the algorithm is not easy to categorize into the local optimal solution in the later stage of optimization, but it has excellent global search ability. Ten test functions were used to evaluate and compare the effectiveness of the original beluga whale optimization algorithm, the improved beluga whale optimization algorithm, and three other prevalent optimization algorithms, focusing on their convergence characteristics. Then, a mechanical analysis was carried out on the bridge-erecting machine girder under real loading conditions. According to the design standard of the main girder of the bridge-erecting machine, under the conditions of meeting the requirements of strength, stiffness and stability, an optimization model was established, and the optimization of the main girder of the bridge-erecting machine was carried out. It was verified that compared with the initial girder weight of the bridge-erecting machine, the optimized girder weight was greatly reduced. The results show that the optimization effect is remarkable and the research has significant practical value.