Power unmanned aerial vehicle shooting waypoint automatic planning method based on reinforcement learning

The invention relates to the technical field of electric power unmanned aerial vehicle shooting, in particular to an electric power unmanned aerial vehicle shooting waypoint automatic planning method based on reinforcement learning, and aims to solve the problems of long shooting time, inaccurate shooting data, low shooting efficiency, increased inspection cost and poor stability caused by the fact that an aerial shooting route is not formulated in advance in an existing electric power unmanned aerial vehicle aerial shooting technology. In order to solve the problems of low data acquisition efficiency and the like, the invention provides the following scheme: the method comprises the following steps: S1, preparing unmanned aerial vehicle shooting equipment; the invention aims to improve data acquisition efficiency and accuracy, improve working efficiency and calculation efficiency by adopting a three-dimensional point cloud technology, automatically identify a shooting target based on historical aerial photog