At present, data and operation-intensive advanced artificial intelligence computing technology is developing rapidly, but power system simulation and analysis platforms are not easy to expand the application of artificial intelligence technology. This paper is based on the large-scale Power System Analysis Software Package (PSASP), and additional extends Non-dominated Sorting Genetic Algorithm (NSGA-II), which is able to implement multi-objective parameters coordination optimization of multiple Power System Stabilizer (PSS) in complex power systems. In the implementation process, there proposes PSASP-MATLAB hybrid simulation technology and designs a multi-objective comprehensive optimization function considering the robustness and stability of multi-machine PSS. Then hybrid programming realizes the interactive connection between PSASP platform transient simulation process and NSGA-II algorithm within MATLAB platform. For complex EPRI36 node examples with multiple machines and multiple operating modes, the PSS global optimization control parameters solved by the proposed technique are better than traditional optimization methods in suppressing oscillations, and have good robustness in multiple operating modes.