Abstract: Based on the standard collaborative optimization and aiming at its inherent flaws such as high sensitivity to the initial guess and slow convergence, this paper combines the collaborative optimization algorithm, the hybrid optimization algorithm, and the dynamic relaxation method to put forward a sectionalized dynamic relaxation collaborative optimization algorithm. The proposed algorithm can be applied in the multi-objective optimization of ship engine room stuctures. Meanwhile, the multi-objective collaborative optimization model of the main ship engine room is established to solve the static and dynamic, multidisciplinary and multi-objective optimization of ship engine room structures. Then, the collaborative optimization is validated through the Isight software to draw the optimal solution. It is shown that the step-by-step dynamic relaxation collaborative optimization algorithm achieves both accuracy and efficiency compared with that based on the genetic algorithm, which provides great references to the multidisciplinary and multi-objective design optimization of more complex structures in real projects.