Abstract:
Fully autonomous Unmanned Aerial Vehicles (UAVs) are becoming extremely popular due to their multi-faceted usage in military, commercial, scientific, recreational, agricultural, and various other applications. One of the key feature necessary to ensure full functional autonomy of a UAV is its motion planning technique. Building motion planning capabilities requires a design framework that must consider i) the dynamics of the vehicle; ii) the representation of uncertain environment; and, iii) kinematic constraints. While most of the existing work fail to include all these features together, this work presents a Motion-Planning framework for UAVs that addresses all of them. The proposed framework employs Model Predictive Control based technique to ensure complete resilience to adverse environmental conditions using two sub-components; namely, an adaptive path planner and a predictive path tracker. The framework is empirically validated for different system models with unknown static obstacles operating in the presence of intermittent disturbance. The intermittent disturbance is modelled by injecting white Gaussian noise. Labview based simulation results on a Quadcopter model, demonstrates the robustness of the proposed framework.