复杂产品的拓扑优化设计可以显著改善材料和节能，有效地降低惯性力和机械振动。本研究选择了一种大吨位液压机作为典型的复杂产品，用以表述这种优化方法。基于可靠性和优化解耦模型与基于教与学优化（TLBO）算法，本文提出了一种可靠性拓扑优化方法。将由板系结构形成的支撑物作为拓扑优化对象，具有轻量化和稳定性好的特点。将某种不确定性下的可靠性优化和结构拓扑优化协同处理。首先，利用有限差分法将优化问题中的不确定性参数修正为确定性参数。然后，将不确定性可靠性分析和拓扑优化的复杂嵌套解耦。最后，利用TLBO算法求解解耦模型。该算法参数少，求解速度快。TLBO算法采用了自适应教学因子，在初始阶段实现了更快的收敛速度，并在后期阶段进行了更精细的搜索。本文给出了一个液压机基板结构的数值实例，验证了该方法的有效性。

The topology optimization design of complex products can significantly improve material and power savings, and reduce inertial forces and mechanical vibrations effectively. In this study, a large-tonnage hydraulic press was chosen as a typically complex product to present the optimization method. We propose a new reliability topology optimization method based on the reliability-and-optimization decoupled model and teaching-learning-based optimization (TLBO) algorithm. The supports formed by the plate structure are considered as topology optimization objects, characterized by light weight and stability. The reliability optimization under certain uncertainties and structural topology optimization are processed collaboratively. First, the uncertain parameters in the optimization problem are modified into deterministic parameters using the finite difference method. Then, the complex nesting of the uncertainty reliability analysis and topology optimization are decoupled. Finally, the decoupled model is solved using the TLBO algorithm, which is characterized by few parameters and a fast solution. The TLBO algorithm is improved with an adaptive teaching factor for faster convergence rates in the initial stage and performing finer searches in the later stages. A numerical example of the hydraulic press base plate structure is presented to underline the effectiveness of the proposed method.