It would be nice if you show your solution so that all can learn
rpi4-mid360-ardupilot-dji,copter in this subject ,In this project, there are four components, and my knowledge of each component is not very in-depth—therefore, this approach was a necessity. First, DJI UAVs provide precise RTK positioning. This data is then fed into ArduPilot via a serial port, through a virtual NMEA protocol GPS implemented on a Raspberry Pi. While this is a simulated GPS, the heading and positioning data it provides are highly accurate due to calibration by the DJI device.
A frustrating issue arises with the combination of the MID-360 and the Raspberry Pi: each time it starts up, its initial azimuth angle is 90 degrees (facing east), which differs from ArduPilot’s default north-facing orientation. To resolve this, we have to start either ArduPilot or the MID-360 first. After that, we rotate the started device to match the initial angle of the other device, and only then start the latter. This logic ensures clarity and consistency in the system. 00000144.BIN
Wow, can you share one or more pictures of your system where the alignment of the FC, MID360 are seen.
I’ve noticed that although both the GPS and the POS have their own timestamps, the synchronization between them feels slightly off. The overall direction is correct, but the details just don’t line up. It’s incredibly frustrating, and I’m not quite sure how to handle it—I have too little experience dealing with these kinds of issues. So far, these have only been phased tests, and things are still somewhat unstable. I want to continue exploring this, and along the way, I’d like to gather suggestions from everyone to improve it.
Show the details as multiple requested
when the rtk fix it auto set ekf orign,but the IMU0 and IMU1 point is not the same ,why ??
I still don’t understand why you post such pictures but not the requested information.
Please show detailed your HW setup where we can see how all your components are placed on the frame.
Also what is an IMU point. IMU data are disabled on your log.
The latidude difference from ORGN0 to ORGN1 is around 2cm.
I am attempting something that seems to have never been tried before. I have briefly described my application scenario and requirements, and conducted extensive testing. However, up until now, I am still continuously improving my methods. The reason I haven’t disclosed my specific steps is that I’m afraid of misleading everyone. My English proficiency and computer programming skills are limited. Many times, I’m not sure if translation software has accurately conveyed my intentions. I worry that my expressions might be ambiguous, and my disclosed steps might limit everyone’s thinking.