Vega-based Dynamic Path Planning For UAV Simulation System Design And Implementation

With the rapid development of aerospace technology, UAV in the modern militarywarfare proves to be more and more important. As the key technology of automaticnavigation, path planning algorithm is playing an increasingly large role in thetheoretical study and practical applications.This paper describes the progress of designing and implementing the UAV pathplanning simulation system using Vega3D visual simulation software. The system hasfollowing functions:1. Flight path planning: the System uses Rapidly-ExploringRandom Tree Algorithm to do online real-time planning and generate a feasible pathbased on the dynamic environment;2. Dynamic Simulation: the UAV flights inaccordance with the path in the simulate system and users can observe UAV flight statusin multi-angle;3. Track assessment: the system uses evaluation module to evaluatedifferent algorithms, making the simulation system to be an online test platform.Meanwhile, for the RRT algorithm uses completely random sampling method, thetree explores very slowly. The paper proposes two methods to improve, one is addingprobability when choosing a sub-target point to reduce the number of useless node ofthe random tree and make the path more direct to the target point. Another is generatingtwo trees which extend from start point and end point respectively to reduce the time offinding path. The result of testing RRT algorithm and the two improved methods showsthat these algorithms have achieved satisfactory results, and they have their ownadvantages in different environments.Finally, this paper summarizes my work and proposes my view on how to improveRRT algorithm and how to perfect the system.