search

QUB-ASL / bzzz

Quadcopter with ESP32 and RaspberryPi
MIT License
7 stars 0 forks source link

[WIP] Kalman filter equations #198

Open alphaville opened 4 months ago

alphaville commented 4 months ago

Main Changes

This is still work in progress. I'm creating this PR to keep everyone in the loop. The objective here is to create a whitepaper on how the kalman filter for altitude estimation. The sensors we want to fuse are (i) the ToF, (ii) the barometer, (iii) the GNSS. They have quite different characterstics (update frequency, precision/accuracy, missing values).

An immediate exercise we need to do is to record sensor data, preferably while flying manually. Do record a video at the same time.

jamie-54 commented 4 months ago

Regards the record data while manually flying what data do we want to record?

Anything else?

jamie-54 commented 4 months ago

sensor characteristics

Sensor Max Update Frequency Accuracy Output Resolution
TeraBee TeraRanger Evo 60m 240Hz 40mm below 14m and 1.5% above 14m 5mm below 14m and 20mm above 14m
Seeed Studio Grove VL53L0X module 50Hz +/- 3% at 5Hz at 1.2m and +/- 5% at 50Hz at 1.2m
BMP180 Digital Barometric Pressure Sensor data sheet
alphaville commented 4 months ago

Regards the record data while manually flying what data do we want to record?

  • ToF
  • Barometer
  • GNSS

Anything else?

That should be sufficient. It would be interesting to record accelerations from the IMU too, if possible. We should look at the missing values too, so let's record unfiltered values.

alphaville commented 3 months ago

@pdavid747, can you update the document in this PR using the notes I shared on Teams the other day? We should also include some simulation results in the document.