In order to improve the crashworthiness of the collision post structure, the optimal cross-section configuration of the collision post was obtained based on the topology optimization method, and the impacting finite element model of the collision post was established in the explicit dynamics software Ls-Dyna. Based on the optimal Latin hypercube experimental design method, a polynomial response surface surrogate model of the specific energy absorption and the peak crush force with respect to the wall thickness of the novel multi-cell collision post was established, the multi-objective optimization problem was conducted using the Non-dominated Sorting Genetic Algorithm-Ⅱ and the Pareto front was obtained. The optimization results show that the total energy and specific energy absorption of the novel multi-cell collision post are improved significantly while meeting the structural lightweight requirements compared with the conventional single-cell structure.