【摘要】 智能网联与自动驾驶技术的突破推动汽车座舱由传统功能集成向高阶智能认知演进。高阶认知座舱通过感知-认知-决策-优化的闭环技术架构，构建以人为中心的智能交互新范式。梳理了高阶认知座舱技术架构，聚焦人车交互关键技术，重点分析大语言模型在优化情境理解适应性及意图识别准确性方面的赋能机制，并探讨舱驾一体化框架下多模态状态感知与协同决策技术对实现人-车-环境三域协同与控制权安全接管的支撑作用，总结了当前面临的技术瓶颈与未来研究趋势。更多还原

【Abstract】 Breakthroughs in intelligent connectivity and autonomous driving technologies are driving automotive cockpits from traditional functional integration toward advanced cognitive intelligence. By adopting a closed-loop framework that links perception, cognition, decision-making, and optimization, the advanced cognitive cockpit establishes a new human-centered paradigm for intelligent interaction. This paper reviews the technical architecture of advanced cognitive cockpits, focusing on the key human-vehicle interaction technologies. It analyzes the mechanisms by which large language models enhance situational understanding and improve the accuracy of intent recognition. Furthermore, it discusses how multimodal state perception and collaborative decision-making technologies within a cockpit-driving integrated framework support human-vehicle-environment coordination and ensure safe control handovers. Finally, the paper summarizes current technical bottlenecks and future research trends.更多还原