Quad-rotor unmanned aerial vehicle suspension air transportation system trajectory planning method based on reinforcement learning

Abstract

The invention discloses a quad-rotor unmanned aerial vehicle suspension air transportation system trajectory planning method based on reinforcement learning. The method comprises the steps that firstly, a performance index evaluation function is constructed; according to the optimization principle, the index function is substituted into a Hamilton-Jacobi-Bellman equation, and then the approximatesolution of the HJB equation is solved by updating the output layer weights of an execution network and an evaluation network in real time, so that the optimal control quantity is obtained. Compared with the prior art, the method has the advantages that the influence caused by unknown disturbance of the flight environment can be effectively suppressed, and trajectory planning and accurate positioncontrol of the unmanned aerial vehicle suspension flight system are realized.

Description

technical field [0001] The invention relates to a four-rotor unmanned aerial vehicle suspended air transport system, in particular to a flight trajectory planning method of a four-rotor unmanned aerial vehicle suspended air transport system. Background technique [0002] At present, the flight trajectory planning method of the four-rotor UAV suspended air transport system is divided into two categories: offline trajectory planning and online trajectory planning according to the stage of trajectory design generation. [0003] For offline trajectory planning, the more commonly used methods in the field of UAV suspended air transport systems include trajectory planning methods based on differential smoothing and trajectory planning methods based on optimal control ideas. Among them, the trajectory planning method based on differential smoothing uses the differential form of the original nonlinear system to map the state space of the original system into a low-dimensional smooth...

AI Technical Summary

Problems solved by technology

[0007] Based on the above research status, it can be seen that the current trajectory planning methods for the quadrotor UAV suspension air transport system still have many deficiencies, for example: 1) Some offline trajectory planning methods need to collect a large amount of flight experiment data and repeatedly iterate the variable information of the system Training and calculation are more complicated; 2) Some trajectory generation strategies require high model accuracy and poor robustness against external disturbances

3) The anti-jamming performance of the online trajectory planning method for some quadrotor UAV suspension air transport systems has not been theoretically proven and experimentally verified in response to unknown disturbances and uncertain factors, so its actual effect remains to be tested

Images