A new modeling approach of flexible baseline for distributed POS

Currently distributed POS(Position and Orientation System) system has been widely used in all kinds of aerial remote sensing equipment, while transfer alignment is just an important link of distributed POS. In traditional transfer alignment model, flexural deflections of wings are ignored, so that flexural deflection of flexible baseline brought by internal and external interference will be introduced into actual system work. Thus new transfer alignment model must be built, and generally there are three approaches: First, H8 filtering algorithm with stronger robustness is adopted rather than wings modeling, however, its filtering precision and convergence rate are lower than that of Kalman filtering; Second, transducer are adopted to detect bending deflections of wings, but with the improvement of POS precision, test data provided by PVDF(Polyvinylidene Fluoride) can no longer meet the requirements of measurement; Third, angle observed quantity of flexural deflections are added into Kalman filter formulation, and motion equation of flexible wings is established to estimate and compensate for deformation angles. This modeling method could conclude the structure of aerial carrier, and in this thesis, aeroelasticity is applied to regard wings as cantilever model of Timoshenko type, and it's also added in to filter equation of transfer alignment to make the modeling more flexible. The simulation results show that estimated accuracy of x direction and y direction is very high, error of estimation is less than 3′; estimated accuracy of z direction is lower, the error of estimation is less than 5′, the error of flexural deflection angle is less then 1′, which shows that installation error angle and flexural deflection could be estimated effectively by using the model established in this thesis.