Risk Assessment and Control Strategies for Cascading Failures in HVDC Receiving-End Systems Considering Multiple Uncertainties

Abstract

This paper proposes a cascading failure blocking and control strategy based on physical time-series rolling assessment and feature matching for high-renewable-penetration HVDC receiving-end systems facing strong transient coupling and dual-sided uncertainties. First, a transient voltage ride-through (VRT) margin evaluation index for renewable energy stations based on multi-binary-table sliding window scanning is constructed to rigorously quantify transient limit-violation events with time-cumulative effects under severe disturbances. Second, a sequential evolution evaluation method considering multiple uncertainties is proposed to identify the evolution paths of multi-stage cascading faults while avoiding state space explosion. Finally, an emergency control architecture of "offline database construction and online efficient matching" integrating K-Means clustering, inverse distance weighting (IDW), local Taylor expansion, and successive linear programming (SLP) is proposed for rapid and accurate control generation. Simulations on the CSEE-VS system show that the strategy suppresses transient reactive surges and eliminates delayed voltage recovery, cutting off early-stage fault paths and significantly enhancing system transient voltage stability.