An Optimized A* Algorithm for Cross-country Path Planning of UGV Based on Raster Image Analysis

In order to solve the path planning problem of Unmanned Ground Vehicle (UGV) in cross-country condition, the influence of different terrain and slope on UGV mobility was studied. The Spectral Angle Mapper (SAM) algorithm was used to classify the terrain of hyperspectral raster data, and the cross-country environment was distinguished according to different terrain. The slope of Digital Elevation Model (DEM) raster data was calculated by surface fitting method. A 3D equivalent simulation environment model was constructed based on terrain and slope data. In the 3D equivalent scene, the estimation function of A* algorithm was optimized, and different movement costs were assigned to different landform and slopes. For the UGV cross-country path planning with known starting point and ending point, the path with the least movement cost was calculated. The planned paths take full account of the influence of different terrain and slope, and passed through the terrain with hard geology and gentle slope. Simulation results show the effectiveness of the proposed algorithm.