`sim.from_position` returns wrong gyroscope readings
tianyeeee opened this issue · 7 comments
I tried using traj_ref, gyro, accel = sim.from_position
to generate a trajectory, but found it seems gyro is not correct.
Hi! Could you be more specific? Maybe provide a code snippet.
Thank you very much for your reply.
I just used your sample code in "examples/ins_gps.ipnb". When I run traj_ref, gyro, accel = sim.from_position(dt, lat, lon, t, h=h, p=p, r=r)
this code, the trajectory is OK, but gyro values seem not correct, because I integrated using np.cumsum(gyro)
it gives very small values. However, according to the trajectory, the vehicle should have made many turns larger than, cumulatively, pi
(180 degree). Is that correct?
In addition, the number of values seems incorrect, because there should be t / dt
sensor values (around 10000), here it only has around 1001 inertial sensor values.
I see. The problem is that the sim.from_position/velocity
were simplified to accept time series with time step equal dt
(that is you need to do the interpolation beforehand). And the example was not updated, unfortunately.
I have to say that this library in a poor state at the moment. I have a much improved version (including GNSS and tightly-coupled INS/GNSS) which I have been developing inside a company, maybe at some point I will be able to transfer the code here.
Meanwhile, it would be great if you could fix the example, at least.
I see. The problem is that the
sim.from_position/velocity
were simplified to accept time series with time step equaldt
(that is you need to do the interpolation beforehand). And the example was not updated, unfortunately.I have to say that this library in a poor state at the moment. I have a much improved version (including GNSS and tightly-coupled INS/GNSS) which I have been developing inside a company, maybe at some point I will be able to transfer the code here.
Meanwhile, it would be great if you could fix the example, at least.
I'll have a try. Thanks a lot.
Hi, I'll try to update this example.