Multi — UAV path planning based on improved neural network

摘要

Aiming at the problem of multi-UAV path planning in three - dimensional environment, a fast path planning method is designed by using the improved neural network algorithm. According to the distance between the UAV and the threat, when the UAV is not in the threat area, the method of taking large step is adopted to achieve the purpose of rapid generation of the path. When the UAV is inside the threat area, the method of taking the adjustable step size is used to achieve the fine search path. The UAV rapid out of the threat is achieved by combining with the dynamic adjustable step size and neural network and using the adaptive learning factors, and using the attack revenue and cost function to establish the revenue function, the strategy should be taken is given through the revenue. The simulation results show that the algorithm proposed in this paper not only guarantees the safety of UAV to bypass the threat, but also the better path is obtained and the best decision-making strategy is found.