<![CDATA[General Background Renewable energy integration requires advanced transmission infrastructures capable of managing long-distance power transfer and fast system disturbances. Specific Background Multi-terminal VSC-HVDC grids offer operational flexibility but face challenges related to coordinated control and ultra-fast DC fault protection. Knowledge Gap Existing studies predominantly address control and protection separately, resulting in limited system-level coordination during dynamic and faulted conditions. Aims This study aims to develop and validate a unified hierarchical artificial intelligence-based framework that integrates control and protection for renewable-integrated MTDC systems. Results A three-layer architecture employing deep reinforcement learning for primary control, an AI-based secondary coordinator, and a hybrid AI-driven fault protection scheme was validated through EMT-based co-simulation of a modified CIGRE B4 MTDC benchmark. The framework achieved significant reductions in settling time, voltage deviation, fault detection latency, and post-fault recovery duration, alongside high fault classification accuracy. Novelty The proposed approach introduces a single, coordinated AI-driven architecture that simultaneously governs normal operation and emergency fault response within MTDC grids. Implications The results demonstrate the feasibility of deploying integrated AI-based control and protection to improve stability, resilience, and operational reliability in future renewable-dominated HVDC networks. Keywords: Artificial Intelligence, VSC-HVDC, Multi-Terminal DC Grids, Fault Protection, Renewable Integration Key Findings Highlights: Integrated AI coordination significantly reduced dynamic settling time and voltage deviations under renewable variability Intelligent fault protection achieved sub-millisecond detection with high classification accuracy System resilience under converter outage conditions improved through adaptive secondary coordination]]>
