Data-driven Modeling based on Expert Knowledge for Transition Mode of Tilt Trirotor UAV

Focusing on the high cost and complex operation in establishing the transition mode model of tilt trirotor unmanned aerial vehicle (UAV), combined with a fuzzy modeling method based on TSK fuzzy model. Firstly, the flight data of UAV are collected through simulation. Secondly, the membership function of TSK fuzzy model is designed according to the expert experience, and constant variable mass and moment of inertia is added to the data set of the neural network for Supervised learning and rapid convergence. Finally, the predictive control law is designed. Based on simulation platform, the off-line model is simulated under a fixed tilt curve which prove the effectiveness of the method. The experimental results show that this method can solve the flight state variables, such as attitude, altitude and speed, more quickly and accurately, and this method greatly reduces the cost of model identification.