Abstract: The penetration of renewable energy generation such as wind power generation and photovoltaics in the power grid is increasing continuously, resulting in serious wideband oscillation problems. Impedance-based stability analysis method is one of the effective research methods for wideband oscillation issues of renewable power integration systems. However, due to factors such as the "black/gray box" problem of commercial renewable power generation units (RPGUs) and the random uncertainty characteristics of operating conditions of renewable power plants (RPPs), there are some difficulties in accurate impedance/admittance modeling of RPPs. Therefore, a wideband impedance/admittance online identification method of RPPs based on combined data-driven and model-driven (physics-informed data-driven) is proposed in this paper. For RPGUs with random and uncertain inputs, the wideband impedance/admittance identification models that are suitable for all operating conditions are established based on the data-driven method. Considering the physical structure model of the RPP, the online wideband impedance/admittance identification model of the RPP is established based on combined data-driven and model-driven. Finally, taking the wind power integration system as an example, the stability of the wind power integration system is evaluated online by using the established admittance identification model. The correctness of online stability analysis is verified by time-domain simulation.