Optimization of Path Planning for Unmanned Aerial Vehicles in Emergency Rescue Based on Convolutional Neural Networks

Unmanned aerial vehicles are of great significance to the rapid and accurate positioning and arrival of victims in disaster relief. The existing UAV route planning is mostly based on human intervention and static map information, which is difficult to meet the needs of rapid response to emergencies. To solve the above problems, convolution neural network (CNN) is proposed to be used in this project. This project will make full use of the advantages of convolution neural network to realize the online recognition of rescue targets, obstacles and dynamic environment. The research results of this project will help to improve the precision and fast navigation of UAV in complex emergency. The research shows that the minimum length of emergency relief with convolution network is 10 km. Using convolution neural network to study the optimal route of UAV is an important progress in emergency and an effective means to improve the rescue capability.