 The proposed method uses a reliability constrained, multi-objective optimization approach to control the negative reaction risk of the auxiliary peer, NRAP, caused by multi-source construction uncertainties and traffic growth. It involves adjusting the pavement thickness to reduce the risk of negative reactions. A sensitivity analysis and a reliability analysis using a generalized regression neural network, GRNN, surrogate model were conducted to demonstrate the significance of the uncertainty level and auxiliary peer negative reactions. The Pareto Front examined the balance between construction costs, driving comfort and specified reliability thresholds. The efficiency and accuracy of the proposed method was validated and a real cable-stayed bridge, and the results showed its advantages in controlling the NRAP. This article was authored by Yinting Bai, Xiaoming Wang, Sudan Wang, and others.