Data acquisition of X-plane’s aircraft through matlab for neural network based identification system

The technological development of a highly maneuver aircraft controller is challenging, as the theoretical foundations are difficult to derive and the experiments for developing those methods are expensive. As conventional PID controller could not be a guarantee to work with the same level of accuracy in the entire operating range, a neural network based controller is proposed due to its excellent ability of self-learning and self-adapting, and it could be used to approximate any nonlinear function with strong robustness and fault-tolerant for the nonlinear characteristics of the plant. As the learning mechanism of the neural networks depends on the accurate data from the aircraft, in this research, those data are taken from X-Plane aircraft simulator. Results show that our developed method could acquire the Cessna aircraft's flight data that could be used as system identification and the development of a control system for the aircraft.