多源异构感知融合SLAM算法研究

打开文本图片集

中图分类号：TP242 文献标识码：A 文章编号：1672-3791（2026）02-0056-03

岚1.451460；2.

Research on Multi-Source Heterogeneous Perception Fusion SLAM Algorithm

TIAN Baohui1LIN Yan1LAN Lan1ZHANG Wentao² HUANG Yan1 1.Henan Collge of Transportation， Zhengzhou，Henan Province， 45146O China; 2.Henan Transportation Development Research Institute Co.，Ltd.， Zhengzhou，Henan Province， 45Oooo China Abstract： This Simultaneous Localization and Mapping （SLAM） algorithm based on multi-source heterogeneous perception fusion has broad application prospects in improving the positioning and mapping accuracy of mobile robots，autonomousvehicles，and other complex environments.The masive fusion data generated byvarious sensors such as LiDAR，visual cameras，and Inertial Measurement Units （IMUs） has an increasing demand for computing power，requiring a fusion SLAMalgorithm that can overcome the limitations ofasingle sensor.The Iterative Closest Point （ICP）algorithm based on Kalman filtering，particle filtering，and improved point cloud matching has shown significant superiority，and the ICP-SLAMalgorithm exhibits higher robustnesin dynamic andoccluded scenes.

Keywords： Multi-source heterogeneous; Sensors; SLAM algorithm; ICP point cloud matching algorithm; Autonomous driving technology

自动驾驶技术凭借自主决策与精准控制能力，在交通出行、物流运输、智慧城市建设等领域展现巨大的应用潜力。（剩余3285字）