基于因子图优化的激光惯性SLAM方法研究OA北大核心CSTPCD

SLAM with the integration of LiDAR and Inertial Measurement Unit is an important means of localization and mapping for autonomous driving in denial environments.It comprises front-and back-end two data processing modules.In current back-end data processing,problems including high accumulative error after long-time running,high calculated load and unsatisfactory robustness in complicated denial environments still exist.To address these issues,this paper proposes a new back-end data processing algorithm adapting to laser inertial odometer based front-end algorithm.Factor graph optimization is applied to build the framework.First,an IMUpre-integration model between laser consecutive key frames is built and employed as the optimization factor for describing the IMUdata in factor graph framework,helping reduce the calculated load of data processing.Second,an efficient loop-closing detection method based on the Scan-Context descriptor is introduced,where three-dimensional spatial structure features of the point cloud are converted into a two-dimensional feature map.The calculation efficiency,therefore,is improved in the premise of keeping an eligible loop-closing detection precision.Finally,with the integration of the front-end odometer information,the whole SLAM model is built using odometer factor,IMUpre-integration factor and loop-closing detection factor,and the optimized position-orientation is solved by nonlinear optimization algorithm.In open-source-dataset and real-vehicle-based tests,the proposed DSC-Algo method exhibits markedly higher global localization precision,better calculating performance and robustness in complicated denial environments compared with other mainstream methods including FAST-LIO2,LIO-SAM and SC-LeGO-LOAM.