Giters

PX4 / flight_review

web application for flight log analysis & review

Home Page:https://logs.px4.io/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Why Step Response for (Roll/Pitch/Yaw) Rate is Zero?

MuhammadBilal1 opened this issue · comments

commented

Describe the bug

Hi,
I am using PX4 (1.14.0) in my drone which is using Pixhawk 6C, but since last few days, I am getting an issue:
The Step Response for (Roll/Pitch/Yaw) Rate is Zero in the flight logs (PID Analysis).
I need this doing the PID Tuning of my drone.

1- Sometimes "Step Response for Roll Rate" at ZERO line
2- Sometimes "Step Response for Pitch Rate" is at ZERO line
3- Sometimes "Step Response for Yaw Rate" is at ZERO line
4- Sometimes "Step Response for Roll/Pitch Rate" is at ZERO line

Why is this happening? I am doing manual PID Tuning, flight behavior seems fine (not 100%, but not shaking or crashing), but response is zero.
I am attaching some of the logs here:

To Reproduce

Performing manual tuning.
Almost all of the flights are done in Position flight mode.
I am using UM982 dual antenna GPS.

Expected behavior

The PID Analysis graphs should show the correct values of the PID for each rate instead of a flat line at 0.

Screenshot / Media

image

image

Flight Log

Roll & Pitch are at 0
https://review.px4.io/plot_app?plots=pid_analysis&log=60815230-a1fb-40fd-a64b-54abfcb1d143

Roll, Pitch & Yaw at 0
https://review.px4.io/plot_app?plots=pid_analysis&log=08c14c20-38ca-404e-aee4-30c531839f56

Pitch at 0
https://review.px4.io/plot_app?plots=pid_analysis&log=f804b7d3-5bf6-49cc-bef6-3473c7e4d8b4

Roll & Pitch are at 0
https://review.px4.io/plot_app?plots=pid_analysis&log=edcf0b09-5ea0-4dce-a5d5-115349aea314

Roll, Pitch & Yaw are available
https://review.px4.io/plot_app?plots=pid_analysis&log=4e796e69-30c9-4199-9063-81c6259ba9f6

Software Version

PX4 (1.14.0)

Flight controller

Pixhawk 6C

Vehicle type

Multicopter

commented

Hi
This is when there's not enough motion. The result is best if you fly in Stabilized/Altitude with fast and full stick changes.

From the log it looks like the MC_YAWRATE_P is a bit too high causing noise on the outputs. I would also set MC_YAWRATE_D to 0.

commented

Hi This is when there's not enough motion. The result is best if you fly in Stabilized/Altitude with fast and full stick changes.

From the log it looks like the MC_YAWRATE_P is a bit too high causing noise on the outputs. I would also set MC_YAWRATE_D to 0.

Hi,
Thank you for your response.
Ok. Is this possible to do the PID Tuning by looking at the following graphs?

  • Roll Angle
  • Roll Angular Rate
  • Pitch Angle
  • Pitch Angular Rate
  • Yaw Angle
  • Yaw Angular Rate

A person told me that we can do it from the above graphs like seeing the difference between the Setpoint & the Estimated lines.

For Example:

  • If the difference between Estimated and Setpoint is greater, then we need to increase the I value to increase the speed of response.
  • If the Estimate isn't reaching the Setpoint, then we should increase the value of P.
  • If the Estimate is over the Setpoint, then we should decrease the value of P for that angle.