Research On Filtering Algorithm Of Unmanned Vehicle Integrated Navigation Based On LG-EKF

In recent years,with the continuous improvement of intelligent level,the unmanned ground vehicle(UGV)industry has achieved rapid development.Compared with the traditional ground vehicle industry,the advantage of UGV is that it can still accurately complete specific tasks without human control,which requires the support of long-term continuous and reliable high-precision navigation and positioning technology.At present,integrated navigation using global satellite navigation system(GNSS)and strapdown inertial navigation system(SINS)is the most mature and effective navigation technology for UGV,but this technology depends on the signal accuracy of GNSS.Therefore,how to maintain the high-precision navigation and positioning ability of UGV has become a technical difficulty in the environment of satellite signal blocking or even refusal.This dissertation studies the continuous high-precision positioning and maintenance of UGV in the urban environment with the risk of satellite signal congestion.The main research contents and innovations are as follows:1)The application of LG-EKF in the field of inertial based integrated navigation is studied.The error state models of LG-EKF,state transformation Kalman filter(ST-EKF)and extended Kalman filter(EKF)are established and compared.A group of short-distance land vehicle experiments in dynamic environment are designed to compare and evaluate the variance retention characteristics of 15 state EKF,ST-EKF and LG-EKF algorithms.2)This dissertation studies the navigation and positioning problem when the GNSS navigation system is running normally or there are few satellites in urban environment.Firstly,an SINS/GNSS tightly integrated navigation method based on LG-EKF is proposed;Then,the configuration scheme of navigation sensors on the land vehicle experimental platform is designed,and the performance parameters of related sensors are introduced;Finally,a four hours long-distance vehicle experiment was designed and carried out in the urban environment,including the challenging environment of satellite signal blocking and less than four observable satellites.Taking the measurement results obtained by the post smoothing algorithm of SINS/differential satellite integrated navigation as the benchmark,comparing the horizontal positioning accuracy difference of SINS/GNSS tightly integrated navigation algorithm under three different filtering frameworks.The experimental results show that under the same experimental conditions,the tightly integrated navigation algorithm based on LG-EKF has the relatively best navigation results,which proves the feasibility of this new filtering method in SINS/GNSS tightly integrated system and the value of further research and engineering practice.3)The navigation and positioning problem of UGV when completing long distance mission under the condition of satellite denial is studied.Firstly,a SINS/ LDV integrated navigation scheme based on LG-EKF is proposed;Then,based on the premise of GNSS-denied,three parts and seven groups of land vehicle experiments under different driving environments are designed,including short-distance land vehicle experiment around the school,high-speed land vehicle experiment around the city and cross regional long-distance land vehicle experiment.The autonomous navigation algorithm proposed in this paper is tested and evaluated in detail.The experimental results show that the navigation algorithm based on LG-EKF has higher positioning accuracy in the high dynamic driving environment,which further proves the improvement of the algorithm compared with the traditional EKF algorithm and high engineering application value.At the same time,in the external environment of non extreme weather,LDV also shows the advantage of measurement accuracy relative to OD.The positioning accuracy of SINS/LDV combination is significantly better than SINS/OD integration.Therefore,SINS/LDV integrated navigation algorithm based on LG-EKF is an effective scheme to realize high-precision vehicle autonomous navigation under the condition of GNSS-denied.4)The problem of high-precision navigation and positioning of UGV in urban environment with the risk of satellite signal blocking is studied.Firstly,considering the different driving situations such as good satellite signal,partial satellite signal blocking and complete satellite rejection,a SINS/GNSS/LDV tightly integrated algorithm based on LG-EKF is proposed;Secondly,the working principle of the smoothing algorithm is analyzed.Then,based on the traditional RTS smoothing algorithm,the idea of sliding window is introduced,and a RTS smoothing technology based on sliding window is proposed to optimize the track of the positioning results of the integrated navigation system and eliminate the outlier jumping points caused by the system sudden failure,so as to obtain higher navigation and positioning accuracy;Finally,aiming at the navigation risk environment that UGV may encounter in the process of driving task,a group of four hours long distance land vehicle experiments are designed.The experimental results show that under the same conditions,the integrated navigation algorithm based on LG-EKF has stronger environmental adaptability and reliability.It further proves the improvement of LG-EKF compared with the traditional filtering framework,and this technology has high engineering application value for UGV seamless navigation.