Recently in the balance of the car, I tried the first-order complementary Angle fusion filtering, the effect is so-so, and parameters have great lag didn't adjust good, and not enough smooth,

And then find the second order online complementary code, didn't see, can you explain the popular point?

Float K2=0.2;//the weight value of accelerometer
Float x1, x2, y1.//computing intermediate variable
Float Angle;//filter output Angle

Void Erjielvbo (float angle_m, float gyro_m)
{
X1=(angle_m - Angle) * (1 - K2) * (1 - K2);
Y1=y1 + x1 x dt;
X2=y1 + 2 * (1 - K2) * (angle_m - angle2) + gyro_m;
Angle Angle + x2=* dt;
}

Angle_m: acceleration and the Angle of
Gyro_m: angular velocity
Dt: integration time


Kalman filtering is too complex, not to consider first,

CodePudding user response:

https://blog.csdn.net/seek97/article/details/81294097? Utm_source=blogkpcl6

Look at this article

CodePudding user response:

reference 1st floor P_zz response:

https://blog.csdn.net/seek97/article/details/81294097? Utm_source=blogkpcl6

Look at this article

First-order complementary I understood, it is the second order of don't understand

CodePudding user response:

http://www.doc88.com/p-2819102704066.html
Look at this article, add a layer of integral

CodePudding user response:

It's not hard to, the basis is better can come out!