Abstract: In view of the unsatisfactory effect of conventional projection pursuit model in high-dimensional nonlinear data processing, this paper introduces the wolf pack algorithm of biological swam intelligence algorithm to optimize the PP model and improve the accuracy of projection pursuit model. The index system of comprehensive benefit assessment for soil and water conservation is established based on criteria level of ecological and social benefits; Five classes of soil and water conservation comprehensive benefits are proposed using principles of quantitative analysis. The optimized model is applied in the comprehensive benefits analysis of soil and water conservation in Yunxian County of Hubei Province during the three periods. The results are perfectly in line with the actual conditions, demonstrating that the established assessment index system and the optimized model are reasonable, the assessment method is proved to be feasible; all of which can be considered as new ideas for the comprehensive assessment of soil and water conservation.