Status Report #2 UAV - Adaptable Fuzzy logic controller

STATUS REPORT #2UAV - ADAPTABLE FUZZY

LOGIC CONTROLLER

Team #9Julian JaramilloCraig CobabeEdward RobinsonChris Pitts

Sep 17, 2012

Create a fuzzy logic controller that utilizes a neural network to control a

wide range of aerial vehicles.

OVERALL NEED UAV’s are better than manned systems for many

tasks. However, UAV’s are hard to integrate because creating attitude feedback controllers is difficult. Milestones Fuzzy library code – 9/28/12

Neural Network integration – 10/26/12

Simulation and real flight testing – 11/12/12

TASKS COMPLETED Successfully simulated fuzzy control

over pitch via HIL Interfaced parameters through mavlink Compiled example code utilizing CUDA

NN Libs Solved some controller issues using

simulation. Flight test after rebuild

Crashed T-Rex 450… Again Replaced broken gears in T-Rex 450

ONGOING/PENDING TASKSOngoing Real flight testing of PI controllers with T-

Rex 450 Coding

Pending Tune Model in RC simulator based on real

flight tests More simlations HIL with fuzzy controller –

newly interfaced parameters

UPCOMING TASKS GUI Programming

Neural Network

Tuning the RC helicopter model

Flight testing of Fuzzy controller

10%

15%

15%BLOCK DIAGRAM

CURRENT SCHEDULE

SOME NOTES NEED VIDEO OF LATEST TEST FLIGHT SO

ICAN EMBED IN SLIDES.