Improvement of Batch Processed TRN Considering MAD Statistics

Abstract

TRN (Terrain Referenced Navigation) is based on the comparison of terrain height measurements with a digital elevation map. The comparison can be carried out either based on MAD (minimum absolute differences) or MSD (minimum square differences). Batch processed TRN can be used to provide horizontal position fixes. However, over flat terrain and water, TRN cannot fix the position of the vehicle, because the candidate profiles which have a MAD value similar to the minimum will appear in the data correlation process. To overcome this drawback, we propose two enhancements. First, a decision is made for whether or not to use batched-processed TRN, depending on the degree of the terrain roughness. Second, the statistics of MAD are used to prevent false fixes due to similar terrain in the area of interest. Simulation results show improvement in the navigation performance of the updated data correlation method.