How to use neural networks toolbox on simulink for controlling a dynamic plant of quadcopter that is using 6DoF Euler Angles Simulink block

Hello all, I have done a model for controlling roll, pitch, yaw and Z of a quadcopter, I did it using the 6DoF Euler Angles Simulink´s Block, and I have solved the controll both, with using fuzzy controllers and also with a PID. Now I am trying to use neural networks for control it, and I have tryed the control blocks (Model-Reference and predictive controller), but I don´t know how to use them for this scenario on a propper way, or how I should redo the plant. Because until now, with the other controllers I pick each of the outputs and I do deacoupled control for each one and join them again. I will be very thanked for any help, thank you. Regards.

[fransisca]

Hi emmanuel I encounter the same problem like you I already build my own block with PID controller, But I am so confused with how to connect with neural network training. Can you kindly share how you solve this problem? Would be a lot grateful about that thank you ！

Answer 1

[MS]

Hi Sir Can you please tell me how you use Fuzzy logic for Control?