Research On Integrity Monitoring Method Of Satellite-based Vehicle Positioning Using Cooperative Weight Estimation

Accurate and reliable vehicle location information is of great significance for the novel Cooperative Intelligent Transportation(C-ITS)in the future.Global Navigation Satellite System(GNSS),as a relatively mature positioning method,has been widely used in road transportation systems.However,in complicated urban observation environments,due to the influence of multipath,non-line-of-sight and fault interference,satellite positioning may fail in satisfying the performance requirement of specific traffic applications.Under the Cooperative Vehicle Infrastructure System(CVIS)scheme,the satellite-based vehicle positioning can break the limitation of relying only on its own information,and realize cooperative weight estimation through information interaction.It will further improve the integrity monitoring performance,thereby ensuring the application support capability.Weight allocation of satellite positioning measurements and tracking adaptation of complex operation environment are concerned in this paper.A integrity monitoring method of satellite-based vehicle positioning using cooperative weight estimation is proposed.Effective matching of satellite observation features and observation weights can be realized through the dynamic,collaborative and customized satellite measurements weighting model.Thus,the abnormal observations can be effectively detected,identified and excluded under complex observation environments,which ensures the superior integrity performance for vehicle positioning.The main research work completed in this thesis includes:(1)The satellite observation feature model is established.By evaluating the indices related to GNSS integrity,the overall architecture of dynamic weight estimation for satellite positioning based on CVIS is designed.(2)An off-line modeling scheme consisting of region division,temporal feature construction and Light GBM model generation is designed.An on-line estimation scheme based on the multi-vehicle collaboration is presented.Based on that,a satellite observation quality modeling strategy and cooperative weight estimation algorithm using RTLGB(Region-Time-Light GBM)are proposed.(3)By introducing the observation information from the collaborative neighborhood vehicles and estimated weight information from the RTLGB model,the scheme for fault detection and identification and weight correction of faulty measurements is proposed to enable a cooperative integrity monitoring method for satellite-based vehicle positioning.(4)An experimental platform for multi-vehicle cooperative positioning is established,with which the performance of the RTLGB model is verified using filed data sets.The performance of the comprehensive multi-vehicle cooperative weighting strategy and the capability of the cooperative integrity monitoring solution are verified through fault injection simulation and practical test.The results of this thesis provide an effective way to address risks of the satellite-based vehicle positioning associated with the complexity and dynamics in urban environments.With the advantages of information interaction and collaboration mechanism in the C-ITS,the satellite-based vehicle positioning can break the limitation of the local observation information condition.With the weight of observation,a tight and effective correlation between satellite positioning and the practical operation environment can be established,thereby optimizing the integrity performance level of satellite-based vehicle positioning,preventing possible failure risks.The proposed solution will provide effective and comprehensive supports for the location-based intelligent transportation system applications.Figure 49,table 18,reference 76.