Abstract: In order to solve the disadvantage of conventional dynamic response optimization method, a hybrid adaptive metamodeling (HAM) with combination of spectral element method and the critical points set method (CPSM) which can process the time-dependent constraints are proposed.An indepth exploration of discrete dynamic response in time domain is considered to transform a system of structural dynamics equations into first-order differential equations by using Bubnov-Galerkin method.The dynamic responses are accurately and efficiently calculated.The corresponding metamodel can be adopted from the alternatives using the adaptive strategy.In the optimization process, the number of potential points is determined by uniform grid, and the local optimization is integrated with HAM, which make the optimization more reliable.The time-dependent constraints method is processed by the critical points set, which is the combination of the critical points and adjacent Gauss-Lobatto-Legendre points.The dynamic model of five dimensions vehicle suspension system is established and optimization design is carried out by HAM and CPSM.The results show that two methods are both effective.However, HAM with combination of spectral element is better than CPSM both in accuracy and efficiency.