Unmanned aerial vehicle data acquisition method based on deep reinforcement learning

The invention discloses an unmanned aerial vehicle data acquisition method based on deep reinforcement learning. The method comprises the following steps: performing deep reinforcement learning on an unmanned aerial vehicle data acquisition function; performing a decision-making process of acquisition function deep reinforcement learning; and carrying out error characteristic analysis and calibration on the collected data. Through deep reinforcement learning, the unmanned aerial vehicle senses a corresponding decision action, analyzes and collects environment information, performs decision control on a collection method by using an artificial intelligence method, realizes data collection of the unmanned aerial vehicle, and reduces the complexity of an algorithm through weight sharing and pooling methods of a deep convolutional neural network. And data features are effectively extracted, and sensor static errors and sensor dynamic errors existing in the acquired data are calibrated in a targeted manner, so tha