Design And Simulation On Autonomous Route Planning For UAV In The Two-dimensional Dynamic Environment

Compared with manned aircraft, unmanned aerial vehicles (UAV) has the advantages of small size, low cost, easy to use and low requirements of flight environment. It has gradually become the focus of media at home and abroad, and it has great application prospects both in civilian and military. Route planning ability is an important symbol of autonomy to UAV. Along with the rise of research and use of UAV, the applications of UAV are gradually expanded which puts forward a number of new requirements to route planning technology.Based on the research of autonomous route planning problem for UAV in the two-dimensional dynamic environment, we propose a new hierarchical planning method. Through the analysis of the relative motion between UAV and dynamic threat, the criterion of hierarchical planning is given. Combined with autoregressive prediction, the heuristic direction with the information of relative angle is designed in rolling window to lead to the final target. In the rolling window, the stream function is utilized for avoiding obstacle. Combined with autoregressive prediction, the prediction of Kalman can obtain a higher accuracy. The path planning based on receding horizon and variable step size can deduce the future battlefield situation and make backtrace path. The heuristic direction with the information of threat can improve the efficiency of the route search. This method not only can realize the UAV autonomous avoidance of dynamic threat and online real-time path planning purposes, but also makes the computation online path planning optimized. It makes UAV with "prospective" and "reasonable". It is great significance to improve the autonomy and intelligence of UAV.Through establishing the model of UAV's surroundings and dynamic threat, a simulation model framework is established. Based on this framework, prediction results and the process of route planning are analyzed by computer simulation. Simulation results show the effectiveness of the method in the two-dimensional dynamic environment, and the evaluation results are satisfactory.