Abstract: In the prestack inversion and prestack seismic attributes analysis，the lack of S?wave information usually leads to a bad influence on petroleum exploration by decreasing accuracy of seismic data interpretation. Hence, Prediction of S?wave velocity by using other methods is more significant. Xu?White model，compared with other methods，comprehensively takes account of the influences of porosity，density and clay content，which is suitable for real sand?mudstone formation characterization with higher predictive accuracy. But the appli?cation of Xu?White model has been limited because the aspect ratio of the inclusion can not be measured in the laboratory. This paper studies the Xu?White model，introduces simulated annealing algorithm and particle swarm optimization（PSO）to seek optimal solution，by which the local minimum solution is avoided during searching process，and the accuracy of S?wave velocity is obviously improved. The case study from real seismic data processing in an area of Dagang shows good results in application by using this method