Fast Path Planning for Unmanned Aerial Vehicles by Self-Correction Based on <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML"><mrow><mi>Q</mi></mrow></math></inline-formula>-Learning

You are here: Homepage > Search

Fast Path Planning for Unmanned Aerial Vehicles by Self-Correction Based on $Q$ -Learning

Wang, Ziyang / Yang, Hongbing / Wu, Qingsong / Zheng, Jiafei

This paper addresses a path planning problem for unmanned aerial vehicles with correcting position errors through correction-point navigation, which requires a rapid response when determining the flight path. A two-layer nested iterative hybrid algorithm based on $Q$ learning is proposed to achieve multiobjective optimization by minimizing path lengths and correction times while reducing the complexity of the algorithm and improving the efficiency of path planning. The lower-layer algorithm uses a $Q$ -learning framework based on the experience playback mechanism and exploration/exploitation mechanism. The rewards and punishments for the $Q$ -table values in the lower-layer algorithm are innovatively managed with the unit “path.” The upper-layer algorithm is based on the Pareto multiobjective algorithm. The Pareto frontier is continuously updated with the solutions from the lower-layer algorithm, which timely provides the lower-layer algorithm feedback. Finally, simulation experiments are conducted to evaluate the effectiveness of the algorithm, and the results show that the proposed algorithm outperforms particle swarm optimization.