We were able to solve it once and it was able to take off and land, but when we tested it in the stadium, it suddenly hit the ground and started to rise again. Then we changed the PIDs and this time the two motors in front gave more thrust and pitched in the opposite direction. We tried very hard but couldn’t fix it.
I’m unconvinced this is an issue that is resolvable via PID tuning. In larger vehicles it is fairly common for vibration to saturate the IMU sensors, resulting in a bad state estimate.
It is entirely possible, if not likely, that the issue is that the vehicle has an incorrect attitude solution. The requested vibration plots would assist in analysis.