LPOS_X and LPOS_Y wrong way around?

pleschinski · December 20, 2016, 5:07pm

Hi everyone,

EDIT : Please ignore this post, I’m an idiot. LPOS is based on a coordinate system that is set where the aircraft started (turned on or armed, I don’t know, but in this situation it doesn’t matter as they are the same), as explained a while ago in this answer to another of my questions. All the graphs make perfect sense now, and nothing is wrong.

I have been going over some flight logs, and found something very confusing… It looks like the LPOS_X and LPOS_Y are switch around. Is my understanding incorrect that LPOS_VX is forward velocity, and LPOS_VY is velocity to the right?
I have attached velocity and position graphs of a flight below, made by px4tools. The aircraft suffered from some serious dutch roll, and ended in a crash. To give you and an idea what the flight looked like, I also attached a graph of roll attitude. You can see it oscillates about the roll axis, getting worse as time goes on.

Now what confuses me, is that the LPOS_VX graph oscillates in time with the roll attitude, not LPOS_VY. Furthermore, it crosses over the x-axis of the graph, and I’m pretty sure my fixed wing never managed a negative forward velocity.
If you look at the position graph, my plane travels -40m forwards, and 116m sideways. I don’t recall that much wind, and if there was, I was flying into the wind, not across it.

I know it’s not an IMU rotation thing in the parameters, because ATT_Roll, ATT_Pitch etc. are all correct.

Z is also confusing because it says I reached a maximum altitude of 2m, but I know I flew at least 4m high. That’s a question for another time, though.