机器人2026，Vol.48Issue(4)：745-767,792,24.DOI:10.13973/j.cnki.robot.250450
大模型与智能机器人的融合:智能感知、导航与操作
Integration of Large Models and Intelligent Robots:Intelligent Perception,Navigation and Manipulation
摘要
Abstract
This survey systematically reviews key advancements in intelligent perception,navigation,and manipulation empowered by large language models and multimodal large models.At the perception level,deep understanding of environmental semantics and geometric attributes is enhanced through multimodal fusion and language-spatial joint reasoning.For navigation,techniques such as chain-of-thought task decomposition and commonsense reasoning enable the parsing of ambiguous instructions and autonomous exploration in unknown environments.In manipulation,the dexterity and adaptability for complex interactive tasks are improved by vision-language-action models coupled with physical commonsense.Research indicates that the introduction of large models drives a paradigm shift in robotics from"perception-driven"to"cognition-driven",significantly enhancing the system capabilities in contextual reasoning and autonomous decision-making.However,core challenges remain,including cross-modal alignment accuracy,real-time performance,safety,reliability,and Sim2Real generalization.A systematic technical reference and development roadmap are provided for building general-purpose,cognition-enhanced intelligent robotic systems.关键词
智能机器人/人工智能/多模态大模型/智能导航/灵巧操作Key words
intelligent robot/artificial intelligence/multimodal large model/intelligent navigation/dexterous manipulation引用本文复制引用
谢远龙,李文龙,张友民,唐剑,周雪峰,郑世祺,杨恺伦,谢胜泉,王书亭..大模型与智能机器人的融合:智能感知、导航与操作[J].机器人,2026,48(4):745-767,792,24.基金项目
国家自然科学基金(52275488,52575573,52205524,52188102) （52275488,52575573,52205524,52188102）
国家重点研发计划(2023YFB4703800). （2023YFB4703800）
