Unmanned aerial vehicle maneuvering target tracking method integrating Kalman filtering and DDQN algorithm

Abstract

The invention provides an unmanned aerial vehicle maneuvering target tracking method integrating Kalman filtering and a DDQN algorithm, and the method comprises the steps: carrying out the precise estimation of the motion state of a target, obtaining the position and speed of the target, taking the state information of an unmanned aerial vehicle as the input of a neural network, taking the acceleration and angular speed of the unmanned aerial vehicle as the motion output, and carrying out the learning through the DDQN algorithm, completing the training of a flight strategy network, and achieving the autonomous tracking decision of the unmanned plane for a maneuvering target. According to the method, the problem of errors caused by direct distance measurement through a sensor in a traditional unmanned aerial vehicle target tracking task is effectively solved, the method has high application value, and the problem of DQN over-estimation in a traditional DQN algorithm is effectively solved.

Description

technical field [0001] The invention relates to the field of control, in particular to a method for tracking a maneuvering target of an unmanned aerial vehicle, which relates to a Kalman filter algorithm and a DDQN algorithm based on deep reinforcement learning in the computer field, and belongs to the application of interdisciplinary methods. Background technique [0002] Unmanned Aerial Vehicle (UAV), as a new type of aviation aircraft equipment, has become a practical and effective development tool in the fields of military, civilian and scientific research, and plays an important role in the upgrading of the aviation industry, the integration of military and civilian technology, and the innovation of industrial efficiency. Role. In practical applications, unmanned aerial vehicles often need to face task scenarios such as cluster cooperative flight or ground target tracking, which have high requirements for manual operation, task allocation and track planning technology. ...

AI Technical Summary

Problems solved by technology

However, this method only utilizes the fitting ability of the neural network, does not provide the learning ability for the UAV, and cannot be applied to a dynamically changing environment.

Patent CN110806759A proposes an aircraft route tracking method based on deep reinforcement learning, by constructing a Markov decision process model, combined with a deep reinforcement learning algorithm to complete the route tracking, but in this method, the aircraft can only fly in sequence according to the provided information Arriving at each mission point and then completing route following, it is impossible to track maneuvering targets with unknown trajectories, which has certain limitations

Images