Skid resistance is vital to road safety. The skid resistance of steel bridge deck pavement (SBDP) is particularly important due to the special operating conditions and unique structure features for bridges. This study aims to propose an effective method to predict the skid resistance of newly built SBDP at the design stage. First, the experiment was designed by Taguchi method to get the skid resistance-related data. Secondly, sensitivity analysis was conducted to evaluate the influence of material design and construction factors on the skid resistance. Finally, a prediction model based on GA-BP neural network (back propagation network optimised by genetic algorithm) was developed. Results show that the mixture design parameters have stronger influences on the skid resistance than the construction factors of SBDP. From this study, the 9-14-1 (Input neurons – Hidden neurons – Output neuron) GA-BP model with the population size of 60 could yield an acceptable prediction result for the skid resistance.

防滑性对道路安全至关重要。由于桥梁特殊的运行条件和独特的结构特点，钢桥面铺装（SBDP）的抗滑性能显得尤为重要。本研究旨在提出一种有效的方法来在设计阶段预测新建 SBDP 的抗滑性。首先，采用田口法设计实验，获取防滑相关数据。其次，进行敏感性分析，评估材料设计和施工因素对防滑性能的影响。最后，开发了一种基于GA-BP神经网络（遗传算法优化的反向传播网络）的预测模型。结果表明，混合料设计参数对防滑性能的影响大于SBDP施工因素。从这项研究中，