Abstract:Addressing frequent sub-synchronous oscillations （SSO） in multi-terminal VSCHVDC integrated offshore wind clusters，a probabilistic assessment method using the Weibull distribution and a Leaky Least Mean Squares algorithm is proposed to explore SSO dynamics in large-capacity direct-drive wind farms. Findings reveal that increased wind speeds shift SSO frequencies from 40 Hz to around 18 Hz，and the cumulative probability of negative damping ratios rises from 14% to 90%. More turbines and multi-terminal integration expand the SSO frequency range and reduce negative damping cumulative probabilities by 1%~8%，offering guidance for large-scale offshore wind integration.