Many wind and photovoltaic power sources are integrated into the grid via grid-following power electronic converters

which may induce small-signal oscillations and threaten the secure operation of power systems. Emerging energy storage systems not only mitigate the volatility of renewable energy but also enhance system stability through grid-forming control. This paper first derives the generalized operational short-circuit ratio (gOSCR) and its critical value for the coordinated planning of grid-following and grid-forming energy storage systems

serving as indicators of small-signal stability. To address the highly nonlinear characteristics of gOSCR and its critical threshold

a machine learning-based surrogate model is developed and embedded into a multi-type energy storage planning model that simultaneously considers the integration of renewable energy. Furthermore

the proposed joint planning model is reformulated as a mixed-integer linear programming (MILP) problem to improve computational efficiency. Finally

simulations on a 4-bus system and the IEEE 39-bus system validate the effectiveness of the proposed method.