Fuzzy Behavior-Based Online Command Generation for Unmanned Aircraft Guidance in Rough Terrain Environments

Abstract

This research is aimed at developing an efficient online path planner for unmanned air vehicle guidance in completely unknown three-dimensional (3D) rough terrain environments. A novel algorithm is proposed that directly incorporates the vehicle dynamics in the guidance strategy. A suitable point mass dynamic model is also developed. The flight path forms gradually as a result of applying the guidance commands to the vehicle dynamics. A key feature of this approach is real-time assessment of terrain characteristics and using this information in the guidance procedure. The problem is considered within a fuzzy behavior–based framework. The guidance algorithm uses acquired information from the onboard sensors and rapidly issues commands that will guide the vehicle safely to an intermediate position within the sensor range. Two behaviors are introduced: go to target and 3D terrain following/terrain avoidance. The issued commands are then integrated with adjustable weighting factors. Simulation results demonstrate a significant enhancement in vehicle autonomy level. Intelligent decision-making capability afforded by this approach allows for autonomous and safe low-level flight in mountainous areas.