摘要: |
Position reconstruction algorithms (PRAs) based on Gaussian techniques converge very slowly, or diverge, for some geometries of ground unit and aircraft location which are within MLS system coverage. On the other hand, algorithms based on Newton-Raphson techniques usually converge very rapidly but impose a significantly greater storage requirement and computational burden on the avionics. This report presents a modified Gaussian algorithm which uses a relaxation factor to achieve rapid convergence for all geometries, and with a computational burden much less than the equivalent Newton-Raphson algorithm. It presents the theoretical foundations of this algorithm and various results showing its effects. It also compares the algorithm's storage and computational burdens against Gaussian, rotated-coordinate Gaussian, and Newton-Raphson equivalent PRAs in the MLS context. |