Star Pattern Matching Methods In Automation Celestial Navigation System For Aerospace Vehicle

Multi task work model and cross-aerospace flight environment in aerospace vehicle have a higher requirement to that navigation system is able to provide the autonomous,continuous and reliable information.Star identification in celestial navigation system is widely applied in the field of position and attitude estimation in the spacecraft and it will be the key technology to realize aerospace vehicle automation navigation in the future.Therefore,single FoV(Filed of View)autonomous star identification method based on the star geometric configuration,star identification method across multiply discontinuous FoVs fusion and star identification method assisted by INS have been researched respectively according to the flight feature and automation navigation requirement in aerospace vehicle.Aimed that traditional single FoV are not able to provide stars vectors to aerospace vehicle in the orbit flight because of the simple star filter strategy,a star identification of star alternative optimization based on the geometric configuration is proposed.According to the packet storage advantages and disadvantages of stars both in star map and catalogue,corresponded in the existing star catalogue optimization method,the star alternative optimization strategy including star magnitude and geometric positions is proposed in the method with the reference to DOP in GPS.This star identification method have a strong robustness against the magnitude error,star vectors measurement error and interfered star error,corresponding to the star catalogue optimization method;and it is able to apply in the orbit flight environment.Aimed that visual stars in single FoV star sensor are limited in quantities and multiply FoVs are discontinuously arranged because of the body constraint of aerospace vehicle,a star identification algorithm across multiply discontinuous FoVs fusion is researched.First,compared with the measurement principle of the continuous FoV,the observation principle across discontinuous FoVs are analyzed and applied in the discontinuous FoVs fusion.Second,on the analysis in discontinuous FoVs fusion,the concept of invisible area and a star catalogue construction method excluding the invisible area are proposed.Finally,.the star identification algorithm flow with two methods above are summarized.This algorithm reasonably deletes the arguments in the invisible area and decrease the scale of catalogue and calculation of the identification.This algorithm can improve the performance of the autonomous identification applied in the aviation high speed flight.Aimed that tracking stars to maintain the star map by the interframe estimation does not adapt the high dynamic flight in aerospace vehicle because of the big error,a star identification algorithm assisted by INS respectively for single and double FoVs are designed.First,the plan of star identification assisted by INS is proposed.Second,on the space relationship analysis of positions and attitudes from INS and boreisight,a boresight estimation method assisted by INS is proposed and IN S errors are regarded to effect the catalogue radium.Second,part catalogue dynamic establishment is proposed with the estimated boresight.Finally,this algorithm is able to deal with errors from INS positions and attitudes in the dynamic environment based on the part catalogue and scale reduction;and the algorithm can provide the principle reference to the aerospace vehicle comprehensive applications.On the basis of star identification principle,the star identification simulation platform for aerospace vehicle is designed,which includes 5 models: flight trace simulation model,GPS/INS integrated navigation model,star map simulation model,variable modes star identification model and display and output model.Through the simulation test under flight trace by the professional software,the availability is verified that star identification methods above apply in the typical flight trace of aerospace vehicle.