With the consumption of the pesticide, it is hard thing to realize full envelop control for quadrotor unmanned aerial vehicle. The parameter variation model of pesticide are constructed firstly. To deal with multi disturbances and system unmodeled parts, a compound model reference control method composed by the adaptive radial basis neural network and state observer is designed. With the selection of the time-invariant linear reference model, the adaptive radial basis neural network is constructed and optimized online to generate time-varying control coefficients. Moreover, the residuals errors are eliminated by the disturbance observer. The cruise simulation tests with different pesticide consumptions verify the compound model reference control method can deal with disturbance torques effectively, and the change of model parameters has little effect on the control performance.
The Model Reference Adaptive Control Method for Pesticide Spraying Quadrotor UAVs
Lect. Notes Electrical Eng.
2021-10-30
10 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch