15 | Combine a gyroscope and accelerometer to measure angles - precisely
ฝัง
- เผยแพร่เมื่อ 2 พ.ค. 2024
- Full code and manual on GitHub: github.com/CarbonAeronautics
In this video, you will learn how you a Kalman filter can combine gyroscope and accelerometer measurements from the MPU-6050 to give accurate roll and pitch angle data to the flight controller.
The purpose of this video series is to learn the basics behind a quadcopter drone and enable you to build one yourself, by dividing this challenging project in several easy-to-understand parts. You use the capable Teensy 4.0 microcontroller together with the easy-to-use Arduino language. - วิทยาศาสตร์และเทคโนโลยี
Great Job. I am also making a quadcopter using a complementary filter. After watching your Kalman filter video, I am thinking of incorporating your approach in my flight controller. Can't wait to see the next videos.
Thanks! Well in the end, a complementary filter is a 'stationary' Kalman filter, meaning that the steady state of the Kalman filter (here 0.005) will be similar to the value you will choose for your complementary filter. So I guess you should get similar results for both.
I also recommend the Madgwick Quaternion update filter :)
you can also use madgwick filter with gradient descent, it will be lighter for the microcontroller and more accurate than kalman
Good idea, didn't hear of this type of filter before, definitely worth trying
Did you do that sir? , please give code if you did@carbonaeronautics
It will be lighter, but for a linear system with Gaussian errors only, the Kalman filter is an optimal solution. You might think of it as a Madgwick filer where the value of "alpha" is determined dynamically rather than being fixed.
I can't thank you enough! You are making my lifelong dream come true.
Great explanation, thank you 👍
This incredible work to share. Thanks so much
Hi! Great Video! Thanks! Can You please share with us the theoritical background of Your Kalman Filter application?
Have you tried reading the data from the MPU6050's onboard Digital Motion Processor? Reading the DMP register data returns a quaternion.
Great Sir now start work on rocket control system ...
we learned a lot from you Sir
thanks for the video, is there any video that just shows what todo with the calibration data?
isn't the MPU-6050 not appropriate for calculating yaw angles?
Or similar kalman filter method could be used for yaw too?
I think there is an issue. When Gyro Integration is happening in 3d, ideally you have rotation matrices multiplication at each step. But your code and logic does not adhere to it. I think it was handled little bit by Kalman filter and if you include that too, error will come down drastically. What say ?
Thank you very much. You should publish. Ready to help!
Thanks
Do you have comparison between Kalman and complimentary filter?
Great Job, quick question though. In the gyro_signals function we are setting Digital Low Filter(0x1A), Sensitivity for Gyro(0x1B) and Sensitivity for Accel(0x1C) and then read the raw values and then do the Kalman Magic. As gyro_signals get called every time we want to read the sensor data it will set the Digital Low Filter(0x1A), Sensitivity for Gyro(0x1B) and Sensitivity for Accel(0x1C) which in my personal opinion is redundant. Is there a reason why we are doing this? In my view Digital Low Filter(0x1A), Sensitivity for Gyro(0x1B) and Sensitivity for Accel(0x1C) can be set as part of the setup. I may be incorrect but thats how I see it.
Thanks! I think you are correct, didn't try it but it sounds logical to put it in the setup part and only call it once. Still room for improvement!
I have tried use your Kalman1D algorithm in my drone application but I have problem with calculated values. It is about 3.5 X lower than angles calculated from gyro or acc. In your application calculated values of kalman angles was similar like form gyro or acc?
does it possible to calculate the velocity using this methode ?
At 3:46 you show Eqn. 4 for updating the predicted state. In this equation the first term on the RHS is given as S(k). Is this correct or should it be S(k-1)?
Hi, I agree with your opinion
For me, the roll and pitch angles are slow to mimic what i am doing with the gyroscope, especially returning to equilibrium. Anything I can tweak in the code to make it more responsive?
This is probably because your hardware is not able to perform the calculations fast enough. The code given uses 4ms as the loop time, and the hardware used (teensy) is fast enough to handle this. And also the IMU output datarate has to fast enough too otherwise the code will endup waiting a long for the data. This will all affect the output.
As a probable solution, first find out the time its currently taking for one iteration to complete. And then adjust the code with this new Time Ts. Things should be more responsive now
Try madgwick & Mahony filter for RPY angles
supercool
How to combine the Magnetometer yaw rotation in kalman filter
What about free fall situation? As I understand in this situation accelerometer will show almost zero vector...
00:03 that glass building and wind mills - looks like Antwerp.
Great video. I may have got a clear understanding on KF finally.
But one thing I must ask, how to get the yaw angle?
you cannot compute the yaw angle with aid of the accelerometer. Therefore, we assume 0 for the yaw angle.
Reading elsewhere it appears important to calibrate the sensor for temperature variation. Per the MPU-6050 spec sheet, the gyro ZRO Variation over temperature (-40C to +85C) is +/- 20 degrees per second. That is a large temperature variation, but it is also a huge error. Your thoughts? Perhaps in a quadcopter the flight duration isn't long enough to become a problem?
How much of that temperature range is applicable?
If raw data change immediately ex 0-90 while the gyro is actually 90 the Kalman should take for a long time til 90 as I saw. How to Fix?
Sir, i use the kalman filter in arduino uno but their is a little delay to changing the angle is that for the low clock ⏰ speed of Atmega328p?
Please reply me sir.
Can you explain to me why std dev of rate is 4 d/s and std dev aggle is 3?thanks
is this a Kalman Filter or an Extended Kalman Filter ?
I am trying to implement this on the rasperry pi pico, but it doesn't work properly, and I cant figure out why
Is kalman really better than complimentary filter ? It'll does the work for me right?... In case of esp32 or nodemcu
If its for a drone the results will be similar for both!
@@carbonaeronautics thanks sir for confirmation
Hii How did you set the iteration length 0.004s i.e the value of G .and how can we change it .which registers do we need to update
The iteration length in the while loop. If you see there is a line while(micros()-loop_timer
Would you explain why you wouldn't use pull up resistors on MPU6050?
Do you mean pullups for the i2C? Are those even really necessary? I've used MPUs before, and I don't have any problem not using those pullups. Have you had problems without those?
can i use this kalman filter code for a gyroscope of my esp32 plane? can the kalman filter handle G force when i try to manuver the plane?
I was also interested in this. I believe this implementation will not work in banked turns based on a test I just ran. I put the unit level in my car, and then drove in a moderate speed turn in a parking lot. The output (serial monitor) showed a 20 degree change in roll. [No my car doesn't sway that much] Similarly on moderate braking, it showed a 20 degree pitch change. Looking at the code, the unreliability of the 6250's gyros is the root problem (I think). We have the 4 * 4 expression [KalmanUncertainty=KalmanUncertainty + 0.004 * 0.004 * 4 * 4;] that is basically saying that the standard deviation error of the gyro is 4 degrees PER SECOND. Then as the gyro estimate diverges from the accelerometer estimate we apply accelerometer corrections. If a turn lasts 1 minute, we are saying the gyro data could be off 4*60 = 240 degrees (on average)! Without some other way of correcting (like knowing speed and rate of turn, thus can calc centripetal force) we need a gyro sensor that is about 20x more trustworthy. I am using a MPU9250, which is better. I'm going to try reducing the 4 to 2 since the MPU9250 is about 2x better and re-test. Still expect it to be unuseable, but will be interesting to see if the change does what I think it will. Note - I'm a complete NOOB in this area, I may be wrong! See video at 1:54
Can we add the yaw angle in a similar way, the only difference being introducing a reference yaw angle to begin with?
You'd use a magnetometer for yaw, which would give you constant accurate results regardless if acceleration.
@@marcelloziglioli8954 oh I see thank you so much
The problem is that the accelerometer cannot tell you the yaw angle. So you could integrate the gyro values, but would not have anything to compare them with -- unless you use an additional sensor such as a magnetometer.
Hi, I'm sorry. I have follow your tutorial since beginning using STM32F103, and seems everything working fine, however, for this lecture 15, it doesn't show any value, the serial monitor literally shows blank, can you please help me
Does measuring the trajectory of a turning plane only require a gyroscope or also an accelerometer? (for trajectory measurement and not that of pitching)
THANKS
It depends. If you have reliable and low-latency information on the speed of the plane and can assume windspeed to be negligible, then just using orientation data from a gyro can give you a reasonably faithful estimate of the trajectory. That being said, should you not have good airspeed data, you probably need accelerometer data to get a trajectory estimate
hey there i just tried this kalmanfilter on my raspberry pipico with mpu6050 and it works really great
thankyou for sharing💫
What a coincidence . I also thinking to do it for RP pico w 🌚
Great to hear it works on the Pi Pico as well!
@@P0K0 great , if you need i could provide the full code
@@pykid1915 Thanks man, for your kind words ! but i would rather code it myself ... If i ran into any problems I'll surely ask .
@@P0K0 🤗
Super!! What about Z axis?
Only the roll and pitch angles are of interest for an angle mode flight controller, because for the yaw direction you generally do not want to return to the initial position, hence rotation rate control in this direction is sufficient
Does any part of the code change if I'm using an Mpu-9250 instead?
I'm using an MPU-9250 with the code, and it seems to work fine on the bench for me. I believe the 9250 is code compatible with the 6050 --- just some of the tolerances are different when I compare the data sheets.
As I had an UNO R3 on hand, I'm trying the code on that. Checking the loop time I do find that the UNO can't keep up. Each step is taking over 4000 ms, which will introduce errors to the result (if I understand correctly). FYI to anyone playing around with this sketch.
when started, my Roll and Pitch value after using Kalman is not 0, but is 30 for roll and 78 for pitch, although I keep MPU lie still
please help me!!!!
I just fix it
how can i stack multiple layers of kalman
Are you Wernher von Braun
or gigabrain?
Need to have a patreon account
ah, da's Antwerpen toch?
42 = 9 + 16