LV Renga is an amazing guy. Because he finds solutions. As soon as he came down, he was right into it. The reason why even the simplest of compass programs given on the amarino website wasn't working on the arduino was because we'd used the analog TX RX pins. Nithin and me were thinking, OK something is wrong in the code, how can we change the code. But, Lv thought something was wrong with what we'd done. So, what he did was just put the TX RX wires into the TX0 RX0 terminals. And, that was what we did constructively today.
Last night, got the pentacopter model worked out.
From now on, my only job is to get one thing done at a time- the simulation will be worked on again and again.QUANTIFY
1. argument 1. The guys want to do adaptive thresholding to get stable hovering. I say use PID loop. why?
1.a. the kk multicopter board used a PI loop for stability.
1.b. The ton of info that we have 'already seen' collectively does not point to using adaptive thresholds for stabilization.
1.c. why try to reinvent the wheel, this is simple, sounds simple, is probably going to give a lot more problems that we can't deal with, there is no blog info, which is pretty much what we've been living on
2. argument on the side of adaptive thresholding.
2.a. no need to use matrices i.e. matrices are tough. kalman filters are much worse.
2.b. we have two months. not much time that. PRACTICALITY.
I have to quit watching these videos for SLAM, UCF CRCV because I need to finish the simulation by the new year.
No comments:
Post a Comment